In this bachelor thesis, I examine how automation of fused filament fabrication (FFF) can be implemented, and what the limitations are for different kinds of automation solutions for FFF. Fused filament fabrication is a 3D-printing technology where a material is extruded through a nozzle, layer by layer, to create a print. The thesis also provides a calculation for the commercial feasibility of small batch rapid production with the implementation of an automation solution for FFF. The approach was a qualitative study containing five interviews, combined with empirical knowledge and data from the additive manufacturing company Svensson 3D. This was complemented with an analysis of which criteria to use when evaluating FFF automation solutions, and a framework for looking at FFF from an operator perspective. To calculate commercial feasibility of automation solutions for FFF, Internal Rate of Return and Payback Time were used. This resulted in six criteria to evaluate solutions for automation of FFF, three evaluations of problems within three solutions for automation of FFF, and a finding showing that small batch rapid production is commercially feasible with automated FFF. Lastly, the thesis contains a discussion regarding what the future is for FFF, and the limitations of the framework presented for evaluating automated FFF systems. Possible promising solutions for automated FFF are presented, together with ideas for how design for additive manufacturing can help shape the future of automated FFF. / I det här kandidatarbetet undersöker jag hur automatisering inom fused filament fabrication (FFF) kan implementeras, och vad begränsningarna är för olika sorters automatiseringslösningar för FFF. Det läggs även fram en uträkning för den kommersiella gångbarheten för small batch rapid production med implementeringen av ett automatiskt FFF-system. Tillvägagångsättet bestod av en kvalitativ studie baserad på fem intervjuer, kombinerad med empirisk kunskap och data från additiva tillverkningsföretaget Svensson 3D. Det här kompletterades med en analys av vilka parametrar som bör användas för att utvärdera lösningar för FFF-automatisering, och ett ramverk där automatiseringslösningarna betraktas ur ett operatörs-perspektiv. För att räkna ut den kommersiella gångbarheten för automatiseringslösningar av FFF användes internränta och återbetalningstid. Det här resulterade i sex parametrar för att utvärdera automatiseringslösningar för FFF, tre utvärderingar av vilka problem som finns i tre existerande automatiseringslösningar, och slutsatsen att small batch rapid production är kommersiellt gångbart för automatiserad FFF. Slutligen innehåller arbetet en diskussion gällande framtiden för FFF och begränsningarna hos det ramverk som presenterades för att utvärdera automatiserade FFF system. Möjliga lovande lösningar för automatiserad FFF presenteras och hur design för additiv tillverkning kan hjälpa till att forma framtiden för automatiserad FFF.

The dielectric antenna is an interesting alternative to a metallic antenna. This is mainlydue to its low manufacturing cost and the possibility to fabricate complex antennageometry with the aid of additive manufacturing (AM). Sophisticated AM technologyprovides new degrees of freedom in shaping the outer and inner geometry of antennas.This feature can be utilized to optimize various properties of antenna, such as itsbandwidth, radiation pattern etc, while maintaining a compact geometry.This master thesis investigates the possibility of improving the bandwidth of acompact dielectric antenna by modifying its geometry. Specifically, dielectricresonator antennas (DRAs) have been considered here. In this connection, twoembedded cylindrical DRAs operating within 8 GHz-17 GHz frequency band havebeen designed and simulated using Ansys HFSS. For the first design (Design-1), abandwidth (corresponding to reflection coefficient ≤ -10dB) of approximately 63%has been obtained and the second design (Design-2) has a bandwidth (correspondingto reflection coefficient ≤ -10dB) of about 57%. However, in terms of radiationcharacteristics, the performance of Design-2 has been found to be superior comparedto Design-1, mainly due to its symmetrical geometry. Furthermore, the two designshave been compared to an existing compact rectangular embedded DRA. It has beenfound that both Design-1 and Design-2 have comparatively wider bandwidth. Withrespect to the radiation characteristics, the performance of the reference antenna andDesign-2 are similar. While, the radiation performance of the reference antenna isfound to be better than Design-1. / Dielektriska antenner är ett intressant alternativ till metalliska diton. Detta beror delspå lägre tillverkningskostnader men också, tack vare additiva tillverkningsmetoder,på grund av möjligheten att använda komplexa geometrier. De senaste årens framsteginom additiv tillverkning har öppnat upp nya möjligheter vid designen av den externaoch den inre geometrin hos dielektriska antenner. Detta kan utnyttjas till att optimeraolika aspekter hos antennen, exempelvis bandbredd och strålningsmönster, utan attpåverka de yttre måtten.Denna avhandling studerar möjligheten att förbättra bandbredden hos dielektriskaresonansantenner (DRA) genom att modifiera deras inre. Två cylindriska DRA:er,verksamma inom 8-17 GHz, har designats och simulerats i Ansys HFSS. Bandbredderom 63 % för Design-1, samt 57 % för Design-2, erhölls. Trots den första designensstörre bandbredd uppvisar Design-2 bättre strålningsegenskaper, främst avseendeantennens strålningsmönster. De simulerade antennerna har också visat sig hastörre bandbredd jämfört med en redan existerande kompakt, inbäddad DRA. Sett tillstrålningsegenskaper är prestandan hos Design-2 jämförbar med referensantennen,medan design ett uppvisar sämre prestanda.

Vid konstruktion av flygplan uppkommer svårigheter att efterlikna de tänktadriftsförhållandena vid vindtunnelprovning av mindre skalmodeller. Dessa svårigheter berorfrämst på att ekonomiska och skalbara begränsningar gör att tillräckligt stora vindtunnlar intekan byggas för att generera de strömningshastigheter som ger samma Reynoldstal som för detänkta driftsförhållandena. För att komma runt detta problem kan så kallade störelementanvändas vilka inducerar en övergång mellan laminär och turbulent strömning i skalmodellensgränsskikt och möjliggör vindtunnelprovning i lägre strömningshastigheter. Detta arbete ämnardärför att ta fram ett kostnadseffektivt och flexibelt designkoncept för störelement ämnade förvindtunnelprovning av en skalmodell. Under arbetets gång undersöktes olika typer avstörelement som sedan togs fram med hjälp av CAD och additiv tillverkning. Dessa testadessedan på en skalmodell i en vindtunnel i syfte att hitta anfallsvinkeln för vilken skalmodellenöverstegrade. Mätdata jämfördes sedan med tidigare simuleringar gjorda för skalmodellen ochslutsatsen som drogs var att de framtagna störelementen hade en otillräcklig förmåga i att fåskalmodellen att överstegra, men att ytterligare modifikation av störelementens höjd och breddskulle göra modellering i CAD och additiv tillverkning till ett attraktivt framtagningskonceptför störelement ämnade för vindtunnelprovning av en skalmodell.

Preface: The research about product development of a prosthetic socket was conducted by two students from Mälardalen University, department of Innovation, Design, and Technology. Background: The most recent public survey shows that an estimated 5 million people in China are amputees, out of which a significantly large portion are below-elbow amputees. Sockets sold to below-elbow amputees are equipped with only two surface electromyography sensors, has low comfortability, has problems with perspiration, and a high weight. The current standard for socket manufacturing has not changed in decades. Research Questions: The following research questions have determined the direction of the research: (1) What measurable factors contribute to a convenient and ergonomic feature design in prosthetic socket from the end-user’s perspective? (2) How can the weight and functionality be improved to achieve a prosthetic socket more suited to the end-user, with respect to the existing prosthetic socket? (3) Which material and manufacturing method is suitable for producing cost-effective and customized prosthetic sockets? Research Method: The research was guided by the 5th edition of Product Design and Development by Ulrich & Eppinger (2012) where the product development process described in five of the six phases from planning to test and refinement were utilized. The data collection and analysis techniques performed in this research was guided by Research Methods for Students, Academics and Professionals by Williamson & Bow (2002). Interviews were conducted with five different stakeholders to find specifications of requirements and concretize subjectivism of what defines quality and ergonomics. Implementation: Currently, below-elbow amputees order sockets from orthopedic clinics. The socket was identified as a product of Ottobock. Investigations were made to find optimal solutions to the specification of requirements. Results: The development of a socket concept was designed for additive manufacturing using a multi-jet fusion printer. Analysis: This concept had significant improvements to parameters: higher grade of customizability, 30 % reduced weight, 48 % cost reduction, a new production workflow with 93,5 % automation, and a 69 % reduction in manual work hours. Conclusions: The data of the research strongly indicate existing potentials in enhancing socket design techniques and outputs by implementation of additive manufacturing processes. This can prove to be beneficial for achieving more competitive prosthetics and associated services. / Förord: Denna forskning om produktutvecklingsprocessen av en armprotes genomfördes av två studenter från Mälardalens universitet, avdelningen för innovation, design och teknik. Bakgrund: Den senaste offentliga undersökningen visar att cirka 5 miljoner människor i Kina är amputerade, varav en betydligt stor del är under-armbågsamputerade. Armproteser som säljs till underarmsamputerade individer är utrustade med endast två yt-elektromyografiska sensorer, har låg komfort, har problem med perspiration och hög vikt. Den nuvarande standarden för armproteser har inte förändrats under årtionden. Forskningsfrågor: Följande forskningsfrågor har bestämt riktningen för forskningen: (1) Vilka mätbara faktorer bidrar till en praktisk och ergonomisk funktionsdesign i underarmsproteser ur slutanvändarens perspektiv? (2) Hur kan vikten och funktionaliteten förbättras för att åstadkomma en underarmsprotes som är bättre anpassad för slutanvändaren med avseende på den befintliga underarmsprotesen? (3) Vilket material och tillverkningsmetod är lämpligt för att producera kostnadseffektiva och anpassade underarmsproteser? Forskningsmetod: Forskningsmetoden styrdes av den femte upplagan av Product Design and Development av Ulrich & Eppinger (2012) där produktutvecklingsprocessen är uppdelad i sex faser. I denna forskning användes de fem första faserna från planering till testning och justering. Tekniker för datainsamling och analys som användes i denna forskning styrdes av Research Methods for Students, Academics and Professionals av Williamson & Bow (2002). Intervjuer genomfördes med fem olika intressenter för att hitta kravspecifikationer och för att konkretisera subjektivitet för vad som definierar kvalitet och ergonomi. Implementering:  Underarmsamputerade individer beställer för närvarande armproteser från ortopediska kliniker. Armprotesen identifierades som en produkt av Ottobock. Undersökningar gjordes för att hitta optimala lösningar för kravspecifikationen. Resultat: Konceptutvecklingen av en armprotes utformades för additiv tillverkning med hjälp av en multi-jet-fusion-skrivare. Analys: Det här konceptet hade betydande förbättringar av parametrar: högre grad av anpassningsbarhet, 30 % minskad vikt, 48 % kostnadsreduktion, ett nytt produktionsflöde med 93,5 % automatisering och en 69 % minskning av manuella arbetstider. Slutsatser: Data från denna forskning indikerar att det finns starkt potential för att förbättra designtekniker och utgångar av underarmsproteser genom implementering av additiva tillverkningsprocesser. Detta kan visa sig vara fördelaktigt för att uppnå mer konkurrenskraftiga proteser och tillhörande tjänster.

Additive manufacturing (AM) has enabled the possibility for the hospitals to become their own implant producers developing implants that are tailored to patient’s anatomy. Despite the enormous potential of custom-made implants there are challenges that complicate the implementation of them into clinical practice. The aim of this research is to (1) identify the main driving forces and barriers for the delivery of custom-made implants; (2) explore staff stakeholder views and practices related to the implementation of AM in surgery for the development of custom-made implants; (3) formulate recommendations on how to cope with the implementation challenges. The research method was an explorative qualitative study consisted of a literature review on the challenges of custom-made implants in clinical applications coupled with the collection and inductive analysis of empirical data. The empirical study was based on ten semi-structured interviews conducted among both domestic and international hospital managers medical doctors and research engineers. The consolidated framework for implementation research (CFIR) was utilized for data collection. Using the five domains of CFIR, the following results were obtained: (1) Characteristics of individuals: Most research participants indicated a positive attitude towards the innovation expressing self-efficacy to its use; (2) Intervention characteristics: Custom-made implants were perceived to have a relative advantage in surgical practice due to their high degree of observability and geometrical adaptability providing increased surgical quality, perfect patient fit and better understanding of pathologies. However, high implementation costs, low degree of trialability and high degree of complexity in the development process were regarded as drawbacks of the innovation; (3) Outer setting: the regulatory uncertainty and lack of reimbursement limit the accessibility of custom-made implants to low income populations; (4) Inner setting: scarcity of resources, staff resistance to change, insufficient management support, communication difficulties, limited access to educational materials and training opportunities as well as lack of time and innovative capacity were regarded by the majority of participants as implementation barriers; (5) Process: central for the success of implementation is the need for a coherent implementation plan and evaluation process as well as the engagement of key stakeholders such as hospital managers, payers, regulatory and implementation advisors. This dissertation proffers a deeper understanding of the implementation issues related to custom-made implants and offers preliminary recommendations on how to cope with implementation impediments through the use of Rogers diffusion of innovation coupled with concepts from the field of organizational change and innovation management including Clayton’s disruptive innovation.

Additive manufacturing with the use of metals have been a steadily increasing field, being able to create products with a higher degree of complexity than traditional processing techniques. SLM is a popular AM process that uses metal powder as feedstock, and one of the key components of this process is the powder rheology. In recent years the use of a powder rheometer has been shown to be a good way of evaluating powder rheology of metal powders used within AM processes, but there is a clear lack of standardised tests and methods. In this study the Compressibility, Permeability, and Aeration test for 316L powders used within SLM processes was evaluated with a FT4 powder rheometer. 15 powders that had undergone printing in SLM processes were studied. This showed that the compressibility test had the best results in differentiating the bad preforming powders, thereafter the Aeration test. The Permeability test wasn’t able to differentiate the bad preforming powders with the settings used. This study demonstrates that some tests with a powder rheometer can evaluate the powder performance in SLM processes, but further research to evaluate the tests and standardise the settings are needed for clearer test results. / Additiv tillverkning med metall är ett område som stadigt ökat i intresse, främst på grund av möjligheten att producera produkter med en mycket högre grad av komplexitet i jämförelse med traditionella processmetoder. SLM är en populär AM process som använder metallpulver som råmaterial, och en av huvudkomponenterna för processen är pulvrets reologi. Under senare år har användningen av en pulver-reometer visat sig ett bra sätt att utvärdera pulver-reologi för metallpulver som används inom AM, men det finns en klar avsaknad av standardiserade test och metoder. I denna studie utvärderas Kompressabilitet, Permeabilitet, och Aerabilitet testen för 316L pulver producerade för SLM processer med en FT4 pulver-reometer. 15 pulver som genomgått SLM printing studerades. Studien visar att kompressabilitets testets utfall bäst överensstämde med det som setts under SLM processen, och bäst urskilde pulvren som fungerat dåligt att printa med, därefter Aerations testet. Permeabilitets testet kunde inte urskilja de sämre pulvren med de inställningarna som användes. Studien demonstrerar att vissa test och index samlade med ett pulver reometer är mer tillförlitliga än andra när det gäller för att utvärdera pulvrets prestanda inom SLM processer, men vidare forskning och studier krävs för att utvärdera testen och standardisera inställningar baserat på pulvret som testas.

State-of-the-art Machine Learning (ML) algorithms show impressive results for a myriad of applications. However, they operate as a sort of a black box: the decisions taken are not human-understandable. There is a need for transparency and interpretability of ML predictions to be wider accepted in society, especially in specific fields such as medicine or finance. Most of the efforts so far have focused on explaining supervised learning. This project aims to use some of these successful explainability algorithms and apply them to Reinforcement Learning (RL). To do so, we explain the actions of a RL agent playing Atari’s Breakout game, using two different explainability algorithms: Shapley Additive Explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME). We successfully implement both algorithms, which yield credible and insightful explanations of the mechanics of the agent. However, we think the final presentation of the results is sub-optimal for the final user, as it is not intuitive at first sight. / De senaste algoritmerna för maskininlärning (ML) visar imponerande resultat för en mängd olika tillämpningar. De fungerar dock som ett slags ”svart låda”: de beslut som fattas är inte begripliga för människor. Det finns ett behov av öppenhet och tolkningsbarhet för ML-prognoser för att de ska bli mer accepterade i samhället, särskilt inom specifika områden som medicin och ekonomi. De flesta insatser hittills har fokuserat på att förklara övervakad inlärning. Syftet med detta projekt är att använda några av dessa framgångsrika algoritmer för att förklara och tillämpa dem på förstärkning lärande (Reinforcement Learning, RL). För att göra detta förklarar vi handlingarna hos en RL-agent som spelar Ataris Breakout-spel med hjälp av två olika förklaringsalgoritmer: Shapley Additive Explanations (SHAP) och Local Interpretable Model-agnostic Explanations (LIME). Vi genomför framgångsrikt båda algoritmerna, som ger trovärdiga och insiktsfulla förklaringar av agentens mekanik. Vi anser dock att den slutliga presentationen av resultaten inte är optimal för slutanvändaren, eftersom den inte är intuitiv vid första anblicken. / Els algoritmes d’aprenentatge automàtic (Machine Learning, ML) d’última generació mostren resultats impressionants per a moltes aplicacions. Tot i això, funcionen com una mena de caixa negra: les decisions preses no són comprensibles per a l’ésser humà. Per tal que les prediccion preses mitjançant ML siguin més acceptades a la societat, especialment en camps específics com la medicina o les finances, cal transparència i interpretabilitat. La majoria dels esforços que s’han fet fins ara s’han centrat a explicar l’aprenentatge supervisat (supervised learning). Aquest projecte pretén utilitzar alguns d’aquests existosos algoritmes d’explicabilitat i aplicar-los a l’aprenentatge per reforç (Reinforcement Learning, RL). Per fer-ho, expliquem les accions d’un agent de RL que juga al joc Breakout d’Atari utilitzant dos algoritmes diferents: explicacions additives de Shapley (SHAP) i explicacions model-agnòstiques localment interpretables (LIME). Hem implementat amb èxit tots dos algoritmes, que produeixen explicacions creïbles i interessants de la mecànica de l’agent. Tanmateix, creiem que la presentació final dels resultats no és òptima per a l’usuari final, ja que no és intuïtiva a primera vista.

Lattice structures have received a lot of attention as cellular materials in recent years because of their outstanding properties, such as high strength-to-weight ratio, heat transfer, energy absorption, and capability of improving noise, vibration and harshness (NVH) behavior. This type of structure received a boost from additive manufacturing (AM) technology, which can fabricate geometries in practically any shape. Due to economic and environmental requirements, lightweight design is increasingly used in automobile and construction equipment applications. NVH behavior is a crucial issue for construction equipment. However, the conventional structures' NVH behavior is mainly decided by the mass, so silence often requires heavy systems, leading to more energy consumption and emission. Therefore, the environmental trends and the resulting economic competition have limited traditional (heavy) solutions to improve NVH behavior and make the lightweight design more difficult. Novel solutions are necessary to light the difficulty and challenge of combining NVH and lightweight requirements. In this research, topology optimization was implemented on a New Articulated Hauler Transmission (NAHT) component to balance lightweight and NVH behavior. The topology- optimized 3D model was filled by a non-homogenous lattice structure with optimal lattice density via size optimization. Lattice structure optimization is one type of topology optimization, and it is the term for describing these procedures. To fabricate the complicated lattice structure, additive manufacturing (or 3D printing) is required (after topology and lattice structure optimization). The new models were analyzed using the finite element method (FEM), and the results of the analysis were compared with those of the original models. After the comparison, positive results were obtained, demonstrating that topology and lattice optimization can be applied in the design of construction equipment components. According to the results, lattice structure optimization can create a reliable lightweight design with good NVH behavior. Furthermore, lattice structure's organization and layout have a significant impact on the overall performance. / Gitterstrukturer har fått mycket uppmärksamhet som cellulära material under de senaste åren på grund av deras enastående egenskaper, t.ex. hög hållfasthet i förhållande till vikt, värmeöverföring, energiabsorption och förmåga att förbättra buller-, vibrations- och bullerskador (NVH-beteende). Denna typ av struktur har fått ett uppsving av tekniken för additiv tillverkning (AM), som kan tillverka geometrier i praktiskt taget vilken form som helst. På grund av ekonomiska och miljömässiga krav används lättviktsdesign i allt större utsträckning inom bilindustrin och byggnadsutrustning. NVH-egenskaperna är en viktig fråga för anläggningsutrustning. De konventionella konstruktionernas NVH-beteende bestäms dock huvudsakligen av massan, vilket innebär att tystnad ofta kräver tunga system, vilket leder till ökad energiförbrukning och större utsläpp. Miljötrenderna och den ekonomiska konkurrens som följer av detta har därför begränsat de traditionella (tunga) lösningarna för att förbättra NVH-egenskaperna och gjort lättviktsdesignen svårare. Nya lösningar är nödvändiga för att lösa svårigheten och utmaningen med att kombinera NVH- och lättviktskrav. I den här forskningen genomfördes topologioptimering på en komponent för en ny ledad transportörtransmission (NAHT) för att balansera lättvikts- och NVH-beteende. Den topologioptimerade 3D-modellen fylldes med en icke-homogen gitterstruktur med optimal gittertäthet via storleksoptimering. Gitterstrukturoptimering är en typ av topologioptimering, och det är termen för att beskriva dessa förfaranden. För att tillverka den komplicerade gitterstrukturen krävs additiv tillverkning (eller 3D-utskrift) (efter topologi- och gitterstrukturoptimering). De nya modellerna analyserades med hjälp av finita elementmetoden (FEM), och resultaten av analysen jämfördes med resultaten av de ursprungliga modellerna. Efter jämförelsen erhölls positiva resultat, vilket visar att optimering av topologi och gitterstruktur kan tillämpas vid utformning av komponenter för byggutrustning. Enligt resultaten kan optimering av gitterstrukturen skapa en tillförlitlig lättviktsdesign med bra NVH-beteende. Dessutom har gitterstrukturens organisering och layout en betydande inverkan på den totala prestandan.

Sizing (hydrophobization) is one of the most important process steps within the added-value chain of about 1/3rd of the worldwide produced paper & board products. Even though sizing with so-called reactive sizing agents, such as alkenyl succinic anhydride (ASA) was implemented in the paper industry decades ago, there is no total clarity yet about the detailed chemical and physical mechanisms that lead to their performance. Previous research was carried out on the role of different factors influencing the sizing performance, such as bonding between ASA and cellulose, ASA hydrolysis, size revision as well as the most important interactions with stock components, process parameters and additives during the paper making process. However, it was not yet possible to develop a holistic model for the explanation of the sizing performance given in real life application. This thesis describes a novel physico-chemical approach to this problem by including results from previous research and combining these with a wide field of own basic research and a newly developed method that allows tracing back the actual localization of ASA within the sheet structure. The carried out measurements and trial sets for the basic field of research served to evaluate the stock and process parameters that most dominantly influence the sizing performance of ASA. Interactions with additives other than retention aids were not taken into account. The results show that parameters, such as the content of secondary fibers, the degree of refining, the water hardness as well as the suspension conductivity, are of highest significance. The sample sets of the trials with the major impacting parameters were additionally analyzed by a newly developed localization method in order to better understand the main influencing factors. This method is based on optical localization of ASA within the sheet structure by confocal white light microscopy. In order to fulfill the requirements at magnification rates of factor 100 optical zoom, it was necessary to improve the contrast between ASA and cellulose. Therefore, ASA was pretreated with an inert red diazo dye, which does not have any impact on neither the sizing nor the handling properties of ASA. Laboratory hand sheets that were sized with dyed ASA, were analyzed by means of their sizing performance in correlation to measurable ASA agglomerations in the sheet structure. The sizing performance was measured by ultrasonic penetration analysis. The agglomeration behavior of ASA was analyzed automatically by multiple random imaging of a sample area of approx. 8650 µm² with a minimum resolution for particles of 500 nm in size. The gained results were interpreted by full factorial design of experiments (DOE). The trials were carried out with ASA dosages between 0% and 0.8% on laboratory hand sheets, made of 80% bleached eucalyptus short fiber kraft pulp and 20% northern bleached softwood kraft pulp, beaten to SR° 30, produced with a RDA sheet former at a base weight of 100 g/m² oven dry. The results show that there is a defined correlation between the ASA dosage, the sizing performance and the number and area of ASA agglomerates to be found in the sheet structure. It was also possible to show that the agglomeration behavior is highly influenced by external factors like furnish composition and process parameters. This enables a new approach to the explanation of sizing performance, by making it possible to not only examine the performance of the sizing agent, but to closely look at the predominant position where it is located in the sheet structure. These results lead to the explanation that the phenomenon of sizing is by far not a pure chemical process but rather a more physical one. Based on the gained findings it was possible so far to optimize the ASA sizing process in industrial-scale by means of ~ 50% less ASA consumption at a steady degree of sizing and improved physical sheet properties.

Sizing (hydrophobization) is one of the most important process steps within the added-value chain of about 1/3rd of the worldwide produced paper & board products. Even though sizing with so-called reactive sizing agents, such as alkenyl succinic anhydride (ASA) was implemented in the paper industry decades ago, there is no total clarity yet about the detailed chemical and physical mechanisms that lead to their performance. Previous research was carried out on the role of different factors influencing the sizing performance, such as bonding between ASA and cellulose, ASA hydrolysis, size revision as well as the most important interactions with stock components, process parameters and additives during the paper making process. However, it was not yet possible to develop a holistic model for the explanation of the sizing performance given in real life application. This thesis describes a novel physico-chemical approach to this problem by including results from previous research and combining these with a wide field of own basic research and a newly developed method that allows tracing back the actual localization of ASA within the sheet structure. The carried out measurements and trial sets for the basic field of research served to evaluate the stock and process parameters that most dominantly influence the sizing performance of ASA. Interactions with additives other than retention aids were not taken into account. The results show that parameters, such as the content of secondary fibers, the degree of refining, the water hardness as well as the suspension conductivity, are of highest significance. The sample sets of the trials with the major impacting parameters were additionally analyzed by a newly developed localization method in order to better understand the main influencing factors. This method is based on optical localization of ASA within the sheet structure by confocal white light microscopy. In order to fulfill the requirements at magnification rates of factor 100 optical zoom, it was necessary to improve the contrast between ASA and cellulose. Therefore, ASA was pretreated with an inert red diazo dye, which does not have any impact on neither the sizing nor the handling properties of ASA. Laboratory hand sheets that were sized with dyed ASA, were analyzed by means of their sizing performance in correlation to measurable ASA agglomerations in the sheet structure. The sizing performance was measured by ultrasonic penetration analysis. The agglomeration behavior of ASA was analyzed automatically by multiple random imaging of a sample area of approx. 8650 µm² with a minimum resolution for particles of 500 nm in size. The gained results were interpreted by full factorial design of experiments (DOE). The trials were carried out with ASA dosages between 0% and 0.8% on laboratory hand sheets, made of 80% bleached eucalyptus short fiber kraft pulp and 20% northern bleached softwood kraft pulp, beaten to SR° 30, produced with a RDA sheet former at a base weight of 100 g/m² oven dry. The results show that there is a defined correlation between the ASA dosage, the sizing performance and the number and area of ASA agglomerates to be found in the sheet structure. It was also possible to show that the agglomeration behavior is highly influenced by external factors like furnish composition and process parameters. This enables a new approach to the explanation of sizing performance, by making it possible to not only examine the performance of the sizing agent, but to closely look at the predominant position where it is located in the sheet structure. These results lead to the explanation that the phenomenon of sizing is by far not a pure chemical process but rather a more physical one. Based on the gained findings it was possible so far to optimize the ASA sizing process in industrial-scale by means of ~ 50% less ASA consumption at a steady degree of sizing and improved physical sheet properties.:Acknowledgment I Abstract III Table of Content V List of Illustrations XI List of Tables XVI List of Formulas XVII List of Abbreviations XVIII 1 Introduction and Problem Description 1 1.1 Initial Situation 1 1.2 Objective 2 2 Theoretical Approach 3 2.1 The Modern Paper & Board Industry on the Example of Germany 3 2.1.1 Raw Materials for the Production of Paper & Board 5 2.2 The Sizing of Paper & Board 8 2.2.1 Introduction to Paper & Board Sizing 8 2.2.2 The Definition of Paper & Board Sizing 10 2.2.3 The Global Markets for Sized Paper & Board Products and Sizing Agents 11 2.2.4 Physical and Chemical Background to the Mechanisms of Surface-Wetting and Penetration 13 2.2.4.1 Surface Wetting 14 2.2.4.2 Liquid Penetration 15 2.2.5 Surface and Internal Sizing 17 2.2.6 Sizing Agents 18 2.2.6.1 Alkenyl Succinic Anhydride (ASA) 19 2.2.6.2 Rosin Sizes 19 2.2.6.3 Alkylketen Dimer (AKD) 23 2.2.6.4 Polymeric Sizing Agents (PSA) 26 2.2.7 Determination of the Sizing Degree (Performance Analysis) 28 2.2.7.1 Cobb Water Absorption 29 2.2.7.2 Contact Angle Measurement 30 2.2.7.3 Penetration Dynamics Analysis 31 2.2.7.4 Further Qualitative Analysis Methods 33 2.2.7.4.1 Ink Stroke 33 2.2.7.4.2 Immersion Test 33 2.2.7.4.3 Floating Test 34 2.2.7.4.4 Hercules Sizing Tester (HST) 34 2.2.8 Sizing Agent Detection (Qualitative Analysis) and Determination of the Sizing Agent Content (Quantitative Analysis) 35 2.2.8.1 Destructive Methods 35 2.2.8.2 Non Destructive Methods 36 2.3 Alkenyl Succinic Anhydride (ASA) 36 2.3.1.1 Chemical Composition and Production of ASA 37 2.3.1.2 Mechanistic Reaction Models 39 2.3.1.3 ASA Application 42 2.3.1.3.1 Emulsification 42 2.3.1.3.2 Dosing 44 2.3.1.4 Mechanistic Steps of ASA Sizing 46 2.3.2 Physico-Chemical Aspects during ASA Sizing 48 2.3.2.1 Reaction Plausibility 48 2.3.2.1.1 Educt-Product Balance / Kinetics 48 2.3.2.1.2 Energetics 51 2.3.2.1.3 Sterics 52 2.3.2.2 Phenomena based on Sizing Agent Mobility 53 2.3.2.2.1 Sizing Agent Orientation 54 2.3.2.2.2 Intra-Molecular Orientation 55 2.3.2.2.3 Sizing Agent Agglomeration 55 2.3.2.2.4 Fugitive Sizing / Sizing Loss / Size Reversion 56 2.3.2.2.5 Sizing Agent Migration 58 2.3.2.2.6 Sizing Reactivation / Sizing Agent Reorientation 59 2.3.3 Causes for Interactions during ASA Sizing 60 2.3.3.1 Process Parameters 61 2.3.3.1.1 Temperature 61 2.3.3.1.2 pH-Value 62 2.3.3.1.3 Water Hardness 63 2.3.3.2 Fiber Types 64 2.3.3.3 Filler Types 65 2.3.3.4 Cationic Additives 66 2.3.3.5 Anionic Additives 67 2.3.3.6 Surface-Active Additives 68 2.4 Limitations of State-of-the-Art ASA-Sizing Analysis 69 2.5 Optical ASA Localization 71 2.5.1 General Background 71 2.5.2 Confocal Microscopy 72 2.5.2.1 Principle 72 2.5.2.2 Features, Advantage and Applicability for Paper-Component Analysis 74 2.5.3 Dying / Staining 75 3 Discussion of Results 77 3.1 Localization of ASA within the Sheet Structure 77 3.1.1 Choice of Dyes 77 3.1.1.1 Dye Type 78 3.1.1.2 Evaluation of Dye/ASA Mixtures 80 3.1.1.2.1 Maximum Soluble Dye Concentration 80 3.1.1.2.2 Thin Layer Chromatography 81 3.1.1.2.3 FTIR-Spectroscopy 82 3.1.1.3 Evaluation of the D-ASA Emulsion 84 3.1.1.4 Paper Chromatography with D-ASA & F-ASA Emulsions 85 3.1.1.5 Evaluation of the D-ASA Emulsion’s Sizing Efficiency 86 3.1.2 The Localization Method 87 3.1.2.1 The Correlation between ASA Distribution and Agglomeration 88 3.1.2.2 Measurement Settings 89 3.1.2.3 Manual Analysis 90 3.1.2.4 Automated Analysis 92 3.1.2.4.1 Automated Localization / Microscopy Measurement 92 3.1.2.4.2 Automated Analysis / Image-Processing 93 3.1.2.5 Result Interpretation and Example Results 96 3.1.2.6 Reproducibility 97 3.1.2.7 Sample Mapping 98 3.1.3 Approaches to Localization-Method Validation 102 3.1.3.1 Raman Spectroscopy 102 3.1.3.2 Confocal Laser Scanning Fluorescent Microscopy 102 3.1.3.3 Decolorization 103 3.2 Factors Impacting the Sizing Behavior of ASA 104 3.2.1 ASA Type 105 3.2.2 Emulsion Parameters 107 3.2.2.1 Hydrolyzed ASA Content 107 3.2.2.2 ASA/Starch Ratio 109 3.2.2.3 Emulsion Age 110 3.2.3 Stock Parameters 111 3.2.3.1 Long Fiber/Short Fiber Ratio 111 3.2.3.2 Furnish Type 112 3.2.3.3 Degree of Refining 114 3.2.3.4 Filler Type/Content 116 3.2.4 Process Parameters 119 3.2.4.1 Temperature 119 3.2.4.2 pH-Value 120 3.2.4.3 Conductivity 122 3.2.4.4 Water Hardness 123 3.2.4.5 Shear Rate 125 3.2.4.6 Dwell Time 127 3.2.4.7 Dosing Position & Dosing Order 128 3.2.4.8 Drying 130 3.2.4.9 Aging 131 3.3 Factors Impacting the Localization Behavior of ASA 132 3.3.1 Degree of Refining 132 3.3.2 Sheet Forming Conductivity 135 3.3.3 Water Hardness 136 3.3.4 Retention Aid (PAM) 137 3.3.5 Contact Curing 138 3.3.6 Accelerated Aging 139 3.4 Main Optimization Approach 141 3.4.1 Optimization of ASA Sizing Performance Characteristics 142 3.4.2 Emulsion Modification 144 3.4.2.1 Lab Trials / RDA Sheet Forming 146 3.4.2.2 TPM Trials 147 3.4.2.3 Industrial-Scale Trials 149 3.4.2.4 Correlation between Sizing Performance Optimization and Agglomeration Behavior on the Example of PAAE 152 3.5 Holistic Approach to Sizing Performance Explanation 154 4 Experimental Approach 157 4.1 Characterization of Methods, Measurements and Chemicals used for the Optical Localization-Analysis of ASA 157 4.1.1 Characterization of used Chemicals 157 4.1.1.1 Preparation of Dyed-ASA Solutions 157 4.1.1.2 Thin Layer Chromatography 157 4.1.1.3 Fourier Transformed Infrared Spectroscopy 157 4.1.1.4 Emulsification of ASA 158 4.1.1.5 Paper Chromatography 159 4.1.1.6 Particle Size Measurement 159 4.1.2 Optical Analysis of ASA Agglomerates 160 4.1.2.1 Microscopy 160 4.1.2.2 Automated Analysis 163 4.1.2.2.1 Adobe Photoshop 163 4.1.2.2.2 Adobe Illustrator 164 4.1.2.3 Confocal Laser Scanning Fluorescent Microscopy 166 4.2 Characterization of Used Standard Methods and Measurements 166 4.2.1 Stock and Paper Properties 166 4.2.1.1 Stock pH, Conductivity and Temperature Measurement 166 4.2.1.2 Dry Content / Consistency Measurement 167 4.2.1.3 Drainability (Schopper-Riegler) Measurement 167 4.2.1.4 Base Weight Measurement 168 4.2.1.5 Ultrasonic Penetration Measurement 168 4.2.1.6 Contact Angle Measurement 169 4.2.1.1 Cobb Measurement 169 4.2.1.2 Air Permeability Measurements 170 4.2.1.3 Tensile Strength Measurements 170 4.2.2 Preparation of Sample Sheets 171 4.2.2.1 Stock Preparation 171 4.2.2.2 Laboratory Refining (Valley Beater) 171 4.2.2.3 RDA Sheet Forming 171 4.2.2.4 Additive Dosing 173 4.2.2.5 Contact Curing 174 4.2.2.6 Hot Air Curing 174 4.2.2.7 Sample Aging 174 4.2.2.8 Preparation of Hydrolyzed ASA 175 4.2.2.9 Trial Paper Machine 175 4.2.2.10 Industrial-Scale Board Machine 177 4.3 Characterization of used Materials 178 4.3.1 Fibers 178 4.3.1.1 Reference Stock System 178 4.3.1.2 OCC Fibers 179 4.3.1.3 DIP Fibers 179 4.3.2 Fillers 180 4.3.3 Chemical Additives 180 4.3.3.1 ASA 180 4.3.3.2 Starches 181 4.3.3.3 Retention Aids 181 4.3.3.4 Poly Aluminum Compounds 181 4.3.3.5 Wet Strength Resin 181 4.3.4 Characterization of used Additives 182 4.3.4.1 Solids Content 182 4.4 Description of Implemented Advanced Data Analysis- and Visualization Methods 183 4.4.1 Design of Experiments (DOE183 4.4.2 Contour Plots 184 4.4.3 Box-Whisker Graphs 185 5 Conclusion 186 6 Outlook for Further Work 191 7 Bibliography 192 Appendix 207 7.1 Localization Method Reproducibility 207 7.2 DOE - Coefficient Lists 208 7.2.1 Trial 3.3.4 – Impact of Retention Aid (PAM) on Agglomeration Behavior and Sizing Performance 208 7.2.2 Trial 3.3.5 – Impact of Contact Curing on Agglomeration Behavior and Sizing Performance 208 7.2.3 Trial 3.3.6 – Impact of Accelerated Aging on Agglomeration Behavior and Sizing Performance 209