International Journal of Engineering Technologies (IJET) Volume: 3 Printed ISSN: 2149-0104 No: 4 e-ISSN: 2149-5262 December 2017 © Istanbul Gelisim University Press, 2017 Certificate Number: 23696 All rights reserved. International Journal of Engineering Technologies is an international peer–reviewed journal and published quarterly. The opinions, thoughts, postulations or proposals within the articles are but reflections of the authors and do not, in any way, represent those of the Istanbul Gelisim University. CORRESPONDENCE and COMMUNICATION: Istanbul Gelisim University Faculty of Engineering and Architecture Cihangir Mah. Şehit P. Onb. Murat Şengöz Sk. No: 8 34315 Avcilar / Istanbul / TURKEY Phone: +90 212 4227020 Ext. 221 Fax: +90 212 4227401 e-Mail: ijet@gelisim.edu.tr Web site: http://ijet.gelisim.edu.tr http://dergipark.gov.tr/ijet Twitter: @IJETJOURNAL Printing and binding: Anka Matbaa Certificate Number: 12328 Phone: +90 212 5659033 - 4800571 E-mail: ankamatbaa@gmail.com ii International Journal of Engineering Technologies (IJET) is included in: International Journal of Engineering Technologies (IJET) is harvested by the following service: Organization URL Starting Feature Date https://www.openaire.eu 2015 Open Access The OpenAIRE2020 Project iii INTERNATIONAL JOURNAL OF ENGINEERING TECHNOLOGIES (IJET) International Peer–Reviewed Journal Volume 3, No 4, December 2017 Printed ISSN: 2149-0104, e-ISSN: 2149-5262 Owner on Behalf of Istanbul Gelisim University Rector Prof. Dr. Burhan AYKAC Editor-in-Chief Prof. Dr. Mustafa BAYRAM Associate Editors Prof. Dr. A. Burak POLAT Assoc. Prof. Dr. Baris SEVIM Asst. Prof. Dr. Ahmet AKTAS Asst. Prof. Dr. Yalcin CEKIC Asst. Prof. Dr. Ali ETEMADI Publication Board Prof. Dr. Mustafa BAYRAM Prof. Dr. Nuri KURUOGLU Prof. Dr. A. Burak POLAT Asst. Prof. Dr. Ahmet AKTAS Asst. Prof. Dr. Yalcin CEKIC Asst. Prof. Dr. Mehmet Akif SENOL Layout Editor Asst. Prof. Dr. Ahmet AKTAS Copyeditor Res. Asst. Mehmet Ali BARISKAN Proofreader Asst. Prof. Dr. Ahmet AKTAS Contributor Ahmet Senol ARMAGAN Cover Design Mustafa FIDAN Tarık Kaan YAGAN iv Editorial Board Professor Abdelghani AISSAOUI, University of Bechar, Algeria Professor Gheorghe-Daniel ANDREESCU, Politehnica University of Timişoara, Romania Associate Professor Juan Ignacio ARRIBAS, Universidad Valladolid, Spain Professor Goce ARSOV, SS Cyril and Methodius University, Macedonia Professor Mustafa BAYRAM, Istanbul Gelisim University, Turkey Associate Professor K. Nur BEKIROGLU, Yildiz Technical University, Turkey Professor Maria CARMEZIM, EST Setúbal/Polytechnic Institute of Setúbal, Portugal Professor Luis COELHO, EST Setúbal/Polytechnic Institute of Setúbal, Portugal Professor Filote CONSTANTIN, Stefan cel Mare University, Romania Professor Mamadou Lamina DOUMBIA, University of Québec at Trois-Rivières, Canada Professor Tsuyoshi HIGUCHI, Nagasaki University, Japan Professor Dan IONEL, Regal Beloit Corp. and University of Wisconsin Milwaukee, United States Professor Luis M. San JOSE-REVUELTA, Universidad de Valladolid, Spain Professor Vladimir KATIC, University of Novi Sad, Serbia Professor Fujio KUROKAWA, Nagasaki University, Japan Professor Salman KURTULAN, Istanbul Technical University, Turkey Professor João MARTINS, University/Institution: FCT/UNL, Portugal Professor Ahmed MASMOUDI, University of Sfax, Tunisia Professor Marija MIROSEVIC, University of Dubrovnik, Croatia Professor Mato MISKOVIC, HEP Group, Croatia Professor Isamu MORIGUCHI, Nagasaki University, Japan Professor Adel NASIRI, University of Wisconsin-Milwaukee, United States Professor Tamara NESTOROVIĆ, Ruhr-Universität Bochum, Germany Professor Nilesh PATEL, Oakland University, United States Professor Victor Fernão PIRES, ESTSetúbal/Polytechnic Institute of Setúbal, Portugal Professor Miguel A. SANZ-BOBI, Comillas Pontifical University /Engineering School, Spain Professor Dragan ŠEŠLIJA, University of Novi Sad, Serbia Professor Branko SKORIC, University of Novi Sad, Serbia Professor Tadashi SUETSUGU, Fukuoka University, Japan Professor Takaharu TAKESHITA, Nagoya Institute of Technology, Japan v Professor Yoshito TANAKA, Nagasaki Institute of Applied Science, Japan Professor Stanimir VALTCHEV, Universidade NOVA de Lisboa, (Portugal) + Burgas Free University, (Bulgaria) Professor Birsen YAZICI, Rensselaer Polytechnic Institute, United States Professor Mohammad ZAMI, King Fahd University of Petroleum and Minerals, Saudi Arabia Associate Professor Lale T. ERGENE, Istanbul Technical University, Turkey Associate Professor Leila PARSA, Rensselaer Polytechnic Institute, United States Associate Professor Yuichiro SHIBATA, Nagasaki University, Japan Associate Professor Kiruba SIVASUBRAMANIAM HARAN, University of Illinois, United States Associate Professor Yilmaz SOZER, University of Akron, United States Associate Professor Mohammad TAHA, Rafik Hariri University (RHU), Lebanon Assistant Professor Kyungnam KO, Jeju National University, Republic of Korea Assistant Professor Hidenori MARUTA, Nagasaki University, Japan Assistant Professor Hulya OBDAN, Istanbul Yildiz Technical University, Turkey Assistant Professor Mehmet Akif SENOL, Istanbul Gelisim University, Turkey Dr. Jorge Guillermo CALDERÓN-GUIZAR, Instituto de Investigaciones Eléctricas, Mexico Dr. Rafael CASTELLANOS-BUSTAMANTE, Instituto de Investigaciones Eléctricas, Mexico Dr. Guray GUVEN, Conductive Technologies Inc., United States Dr. Tuncay KAMAS, Eskişehir Osmangazi University, Turkey Dr. Nobumasa MATSUI, Faculty of Engineering, Nagasaki Institute of Applied Science, Nagasaki, Japan Dr. Cristea MIRON, Politehnica University in Bucharest, Romania Dr. Hiroyuki OSUGA, Mitsubishi Electric Corporation, Japan Dr. Youcef SOUFI, University of Tébessa, Algeria Dr. Hector ZELAYA, ABB Corporate Research, Sweden vi From the Editor Dear Colleagues, On behalf of the editorial board of International Journal of Engineering Technologies (IJET), I would like to share our happiness to publish the twelfth issue of IJET. My special thanks are for members of editorial board, publication board, editorial team, referees, authors and other technical staff. Please find the twelfth issue of International Journal of Engineering Technologies at http://ijet.gelisim.edu.tr or http://dergipark.gov.tr/ijet. We invite you to review the Table of Contents by visiting our web site and review articles and items of interest. IJET will continue to publish high level scientific research papers in the field of Engineering Technologies as an international peer-reviewed scientific and academic journal of Istanbul Gelisim University. Thanks for your continuing interest in our work, Professor Mustafa BAYRAM Istanbul Gelisim University mbayram@gelisim.edu.tr -------------------------------------------- http://ijet.gelisim.edu.tr http://dergipark.gov.tr/ijet Printed ISSN: 2149-0104 e-ISSN: 2149-5262 vii viii Table of Contents Page From the Editor vii Table of Contents ix  A Multi-Objective Approach For The In-Plant Milk-Run in a Furniture Factory / 186-189 Erhan Baran  Nonlinear Analysis of a Multiple Story Building Under Uniform Blast Loading / 190-195 Namata Saidou Sabiou, Ali Koçak  The Investigation of Wear Behaviour of Dual Phase Steel DP600 / 196-201 Yasin Aygül, İsmail Ovalı  Binding Energy and Stability Calculations on Hydrogenated Forms of Substituted Carbazoles as Hydrogen Storage Materials / 202-206 Mustafa Karakaya, Fatih Ucun  A New Multilevel Inverter Based Parallel Active Power Filter / 207-212 Korhan Karaarslan, Birol Arifoglu, Ersoy Beser, Sabri Camur  Effects of Arc Voltage and Welding Current on the Arc Length of Tungsten Inert Gas Welding (TIG) / 213-221 Ikpe Aniekan E., Owunna Ikechukwu, Ememobong Ikpe  Interactive Fuzzy Decision Making Algorithm for Two Level Linear Fractional Programming Problems / 222-229 Hasan Dalman  Innovative Design for A Ball Worm Gear Mechanism / 230-234 Sait Koçak ix International Journal of Engineering Technologies, IJET e-Mail: ijet@gelisim.edu.tr Web site: http://ijet.gelisim.edu.tr http://dergipark.gov.tr/ijet Twitter: @IJETJOURNAL x INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Erhan Baran, Vol.3, No.4, 2017 A Multi-Objective Approach for The In-Plant Milk- Run in A Furniture Factory Erhan Baran*‡ *Industrial Engineering Department, Faculty of Engineering, Hitit University, 19030 (erhanbaran@hitit.edu.tr) ‡ Erhan Baran, Industrial Engineering Department, Hitit University, Tel: +90 364 227 4533, Fax: +90 364 227 4535,erhanbaran@hitit.edu.tr Received: 18.07.2017 Accepted: 13.12.2017 Abstract- Today, facility optimization is one of the most important decision for the firms while establishing a new factory. It is necessary to carry work-in-process products between the production lines in firms. Many firms use trains to carry these items. In this study, an application is made in a furniture manufacturing factory. Firstly, the routes which the train can use in the firm are created. The alternative routes are compared with each other. The routes contain all of the production lines in the factory. Our aim is to route the train with minimum distance and minimum amount of cargo carrying. In this study, one train is used in the firm. It is also necessary to use minimum train for carrying the work-in-process products. The optimal route is determined by using some of the methods in the literature for multi-objective models. Keywords Facility logistic, In-plant Milk-Run, Multi-Objective Programming, Vehicle Routing Problem, Tow Train. 1. Introduction 2. Literature Review In this study, one of the leading furniture suppliers of the It is necessary to design the optimal route in a facility. The world is investigating a cargo transportation route that can be sequence and the time is very important. In this study, multi- integrated with the production line and the loading of these objective approach is used for vehicle routing for a firm. In the trains. It is aimed to develop this process with a systematic literature, it is called in-plant milk-run when routing is planned viewpoint. In the system studied, it is aimed to optimize the in the firm. In Du et al.’s study, there is a milk-run application transportation of the materials in the production line between between suppliers and customers. Best fit algorithm and 2- the units in the factory. Today, it is very important to minimize exchange algorithm is used for vehicle-dispatching [1]. In the time loss for factories. In order to be able to compete with Jafari-Eskandari et al.’s study, the milk-run between the the big companies in the sector, it is always necessary to carry supplier and the manufacturer has been studied. A model has out the best works. In this context, it is very important that the been developed that takes account of the advantages and trains used in cargo transportation are used regularly on a disadvantages of this method. First, a complex mathematical certain route in a systematic manner with a minimum waiting model has been developed. Then the robust optimization time between units. In this study, appropriate routes are approach is used [2]. In Sadjadi et al.’s study, the milk-run is determined for trains owned by the company, and then a route between the suppliers. A mixed number model has been to reach the whole point was determined with minimum established. However, a meta-intuition was needed because distance in the most appropriate way. is necessary to design the computation time was too long. Then a genetic algorithm the optimal route in a facility. The sequence and the time is was used [3]. In Nemoto et al.’s study, the milk-run is between very important. Next section literature review about this study the suppliers. The milk-run application of an automobile is given. And in other sections, an application is made and the company in Thailand is explained [4]. In Kovac’s study, the solution is compared with the current situation of the facility. milk-run is in stock within the firm. A mixed integer model Optimal route means, carrying the items with minimum has been developed. Solved with ILOG CPLEX [5]. In Brar et distance and minimum time. al.’s study, the literature review for the milk-run is made [6]. In Kilic et al’s study, in-plant milk-run is made. 0-1 mixed 186 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Erhan Baran, Vol.3, No.4, 2017 integer mathematical model has been developed [7]. In Arvidson’s study, the milk-run is between distribution center Table 1. Results for the weighted sum method and customers. During the milk-run application, different routes (clockwise, counter clockwise) were compared and it S olutions Objective (F1,F2) Route was investigated which is better in terms of fuel consumption 1 -8 520,350 1-5-3-4-6-2-1 and time [8]. In Gyulai et al.’s study, a literature review is made for vehicle routing and milk-run [9]. In Ma et al.’s study, 9 -16 450,390 1-5-3-4-6-2-1 the milk-run is between the suppliers. A mutated ant colony algorithm is used [10]. In Hosseini et al.’s study, the milk-run 16-20 420,470 1-5-4-3-2-6-1 is between the suppliers. The harmony search and annealing 21 420,550 1-6-2-3-4-5-1 simulation algorithm is used [11]. In Hanson et al.’s study, the in-plant milk-run is made. A study has been made on how 5. Mathematical Model effective the unit load is in the factory supply [12]. In Novaes et al.’s study, the milk-run is between customers. The dynamic In this study, the mathematical model has two objectives. tooling procedure has been developed. Genetic algorithm is These are; minimizing of total carrying distance and used [13]. In Lin et al.’s study, the milk-run is between minimizing the amount of cargo carrying. The second suppliers. Greedy has developed a two-step heuristic objective means that some packages contain more than the algorithm, which is an intuitive and tabu search algorithm target cell needs. So only demand of the cells can be carried [14]. In Lin et al.’s study, the milk-run is between the by the train. The mathematical model for this problem is given suppliers. A two-step heuristic algorithm, greedy heuristic and below. annealing simulation algorithm, has been developed [15]. In Klenk et al.’s study, in-plant milk-run is made. Various 5.1. Notations strategies have been developed in order to make the ship i,j: Transportation node index, i=1,2,3,…,6 and j=1,2,3,…,6 better, and 4 different algorithms have been developed for cell nodes and i,j=1 is delivery center or depot node. these strategies [16]. In Korytkowski et al.’s study, the in-plant milk-run is made. A milk-run is simulation application was dij: Carrying distance from node i to node j. carried out in an assembly line [17]. In this study, an in-plant R: A set of intermediate nodes. milk-run application is made by using multi-objective programming in. The difference from the literature is to obtain ui: It has any real value of node i. the route under these objectives. lij: Carrying amount of item from the way i to node j. 3. In-Plant Milk-run 5.2. Objective Functions min 𝐹1 = ∑6 6𝑖=1∑𝑗=1𝑑𝑖𝑗𝑥𝑖𝑗 (1) There are many types of problems in the literature for carrying the items from somewhere to the target place. In min 𝐹2 = ∑6𝑖=1∑ 6 𝑗=1 𝑙𝑖𝑗𝑥𝑖𝑗 (2) material flow system, items are transferred to the workstations from the warehouse or several supermarkets. And this transfer 5.3. Constraints can be accomplished using tugger trains or forklifts. These 𝑁∑ 𝑘𝑖=0 𝑥𝑖𝑗 = 1 Ɐj=1,2,….,6 (3) trains show that transferring workstations in one trip of the train. This is called “in-plant milk-run” which will decrease 𝑁∑ 𝑘 𝑗=0 𝑥𝑖𝑗 = 1 Ɐi=1,2,….,6 (4) the transportation costs [18]. This method is different from the other milk-run application because of the train’s capacity and 𝑢𝑖 − 𝑢𝑗 + 𝑛. 𝑥𝑖𝑗 ≤ 𝑛 − 1 Ɐi≠j ϵ R (5) the demand patterns. Just in time is very important for these The equations (1) and (2) are the objective functions of problems. the mathematical model. (3) and (4) constraints are pacified to visit each node at least once. Equation (5) denotes the sub-tour 4. Application elimination constraint. This type of sub-tour elimination constraints, using Gomory cutting planes approach developed In this study, an application is made in a furniture by Miller, Tucker and Zemlin in 1960[19]. These constraints manufacturing factory. In this factory, there is a problem for took the tour of the state model to inhibit the formation of carrying the items from one cell to another cell in the factory. approximately (n2-3n+2) lead to the addition of one constraint The tugger train is used to carry the items. In this study an [20]. In particular, the use of this Miller, Tucker and Zemlin optimal route for the train is obtained by using multi-objective constraints in oversized model is much easier and routing mathematical programming. To solve the multi-objective problems is one of the constraints to improved well-received problems, there are many methods presented in the literature. tour blocking. In this study, €-Constraint Scalarization Method, Weighted- Sum Scalarization and The Augmented Weighted 6. Computational Results Thcebycheff Scalarization Method is compared for this problem. The best of these method is used. There are 6 cells In this study, there are two objective functions. Firstly, in the factory in this study. The demands of these cells are given below in Table 1. combining these objectives process is made. Three of leading methods in the literature applied. These methods are; The 187 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Erhan Baran, Vol.3, No.4, 2017 Weighted-Sum Scalarization Method, The €-Constraint References Scalarization Method and The Augmented Weighted Tchebycheff Scalarization Method. For the weighted-sum [1] Timon Du, F.K. Wang, Pu-Yun Lu, “A real-time vehicle- scalarization method, the mathematical model is solved and dispatching system for consolidating milk-runs”, the results are given in Table 1. Transportation Research Part E: Logistics and Transportation Review, 43, 565-577, 2007. Second method to combine the objectives is the epsilon- [2] M. Jafari-Eskandari, S.J. Sadjadi, M.S. Jabalameli, A. constraint method. In the epsilon-constraint method, one of Bozorgi-Amiri, “A robust optimization approach for the the objective function is written as a constraint. According to milk-run problem (an auto industry supply chain case this situation, objective function and constraints have to be industry)”, Computers & Industrial Engineering, 6-9 July written below. Epsilon values are changing between 350-430. 2009. The results for the epsilon-constraint method is given in Table [3] S.J. Sadjadi, M. Jafari, T. Amini, “A new mathematical 2. modelling and a genetic algorithm search for milk-run problem (an auto industry supply chain case study)”, The Table 2. Results for the epsilon-constraint method International Journal of Advanced Manufacturing Technology, 44, 194, 2009. Solutions Objective (F1,F2) Route [4] Nemoto, T., Hayashi K., Hashimoto M., “Milk-run 1-8 420,550 1-6-2-3-4-5-1 logistics by Japanese aotumobile manufactures in Thailand”, Procedia-Social and Behavioral Sciences, 2(3), 9-16 420,470 1-5-4-3-2-6-1 5980-5989, 2010. 17-19 420,550 1-6-2-3-4-5-1 [5] Kovacs, A., “Optimizing the storage assignment in a warehouse served by milk-run logistics”, International 20-21 420,470 1-5-4-3-2-6-1 Journal of Production Economics, 133(1), 312-318, 2011. [6] Brar, G.S., Saini, G., “Milk-run logistics: literature review The augmented weighted tchebycheff function method is and directions”, Proceedings of the World Congress on based on the minimization of distance from ideal point. The Engineering, 1, WCE 2011, July 6-8 2011. results for the augmented weighted tchebycheff function [7] Kilic, H.S., Durmusoglu, M.B., Baskak, M., method is given in Table3. “Classification and modelling for in-plant milk-run distribution systems”, The International Journal of Advanced Manufacturing Technology, 62(9-12), 1135- Table 3. Results for the augmented weighted tchebycheff function method 1146, 2012. [8] Arvidsson, N., “The milk-run revisited: a load factor Solutions Objective (F1,F2) Route paradox with economic and environmental implications for urban freight transport”, Transportation Research Part 1-3 520,350 1-5-3-4-6-2-1 A: Policy and Practice, 51, 56-62, 2013. 4 490,380 1-5-2-6-4-3-1 [9] Gyulai, D., Pfeiffer A., Sobottka, T., Vancza, J., “Milk-run vehicle routing problem approach for shop-floor 5-11 450,390 1-5-6-4-3-2-1 logistics”, Procedia CIRP, 7, 127-132, 2013. 12-21 420,470 1-5-4-3-2-6-1 [10] Ma, J., Sun, G., “Mutation ant colony algorithm of milk-run vehicle routing problem with fastest completion 6. Conclusion time based on dynamic optimization”, Discrete Dynamics in Nature and Society, 2013. In this study, a multi-objective mathematical is created for the in-plant milk-run problem. The objectives of the [11] Hosseini S.D., Shirazi, M.A., Karimi, B., “Cross- problem are combined by three of leading methods in the docking and mil-run logistics in a consolidation network: literature. Then it is examined which method is more a hybrid of harmony search and simulated annealing approach”, Journal of Manufacturing Systems, 33(4), 567- effective. According to the results that calculated by GAMS 577, 2014. 24.7.1 program, the most efficient scalarization method for the in-plant milk-run problem is ‘Augmented Weighted [12] Hanson R., Finnsgard, C., “Impact of unit load size on in-plant materials supply efficiency”, International Tchebycheff Scalarization Method.’. It has the pare to Journal of Production Economics, 133(1), 312-318, 2011. efficient point more in many place number of these places are higher than the other methods have. The optimal point for both [13] Novaes, A.G.N., Bez, E.T., Burin, P.J., Aragao Jr., objective function is (F =420, F =470). And also optimal D.P., “Dynamic milk-run OEM operations in over-1 2 congested traffic conditions”, Computers & Industrial route for this in-plant milk-run is determined. The route must Engineering, 88, 326-340, 2015. be 1-5-4-3-2-6-1. 188 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Erhan Baran, Vol.3, No.4, 2017 [14] Lin, Y., Xu, T., Bian, Z., “A two-pahse heuristic [17] Korytkowski, P., Karkoszka,R., “Simulation-based algorithm for the common frequency routing problem with efficiency analysis of an in-plant milk-run operator under vehicle type choice in the milk-run”, Mathematical disturbances”, The International Journal of Advanced Problems in Engineering, 2015. Manufacturing Technology, 82(5-8), 827-837, 2016. [15] Lin, Y., Bian, Z, Shuijing, S., Xu, T., “A two-stage [18] Alnahal, M., “In-plant milk-run”, Transportsysteme simulated annealing algorithm for the many-to-many milk- Und, Duisburg, Essen University, 2014. run routing problem with pipeline inventory cost”, [19] Bellmore, M., Nemhauser, G.L., “The travelling Mathematical Problems in Engineering, 2015. salesman problem: a survey”, Operations Research, 16(3), [16] Klenk, E., Galka, S., Günthner, W.A., “Operating 538-558, 1968. strategies for in-plant milk-run systems”, IFAC-Papers On [20] Kulkarni, R.V., Bhave, B.R., “Integer programming Line, 48(3), 1882-1887, 2015. formulations of vehicle routing problems”, European Journal of Operational Research, 20(1), 58-67, 1985. 189 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 Nonlinear Analysis of a Multiple Story Building Under Uniform Blast Loading Namata Saidou Sabiou*, Ali Koçak**‡ * Department of Civil Engineering, Yıldız Technical University, 34210 Istanbul, Turkey ** Department of Civil Engineering, Yıldız Technical University, 34210 Istanbul, Turkey (sabiousaidounamata@yahoo.fr, akocak@yilidiz.edu.tr) ‡ Namata Saidou SABIOU; Ali KOÇAK, Postal address, Tel: +90 (212) 383 5217, Fax: +90 (212) 383 5102, namata.saidou@std.yildiz.edu.tr Received: 22.07.2017 Accepted: 13.12.2017 Abstract- The design of structures to critical can be heavily modified by explosions or shocks effects; in reality, intentional disruptions, blasts, or impacts have unfortunately become part of the possible load scenarios that could act on construction structures during their lifetime. The use of vehicle bombs to attack city centers or bombs intentionally put on construction facilities has been a feature of campaigns by terrorist organizations around the world. A bomb explosion within or immediately nearby a building can cause catastrophic damage on the building's external and internal structural frames, collapsing of walls, blowing out of large expanses of windows, and shutting down of critical life-safety systems. Loss of life and injuries to occupants can result from many causes, including direct blast-effects, structural collapse, debris impact, fire, and smoke. The response of reinforced concrete structures subjected to blast loads can be studied numerically using commercial finite element programs. This paper, hence, to get the structural response of structural elements within a 6 storeys building, which is composed a concrete frame a simulation using the finite element programs ABAQUS. Keywords: Blast, load, structure, finite element, building. 1. Introduction from researchers over the years. Magnusson and Hallgren [3], in their study revealed that concrete beams show an A number of analytical and numerical models have been increased load capacity for blast loading relative to static published in the international literature using software’s loading. Luccioni and Luege [4], by analysing the for Scientifics and engineering designs. Blast loads have behaviour of concrete pavements subjected to blast loads received more attention in recent years because of produced by the detonation of high explosives, performed accidental or intentional events (terrorist attacks) affected the tests to assess the concrete pavement slab under blast important structures. Traditionally, little research on the loads, as a result, the maximum vertical displacement blast resistance of structures has appeared in the open obtained with the finite element model was significantly literature. Enrin et al. [1] and Ishikawa et al. [2] , Studied lower than the one measured experimentally for the third the performance of prestressed concrete beams when explosion using a 12.5 kg charge. subjected to impact loading, bounded and unbounded However, due to the high cost, it is almost impossible to prestressed beams were tested experimentally and investigate the response of the multi- storey buildings against analytically. It was found that, while static loading blast loads with full-scale experimental tests. The analysis resulted in failure of the compression concrete for both the and design of structures subjected to blast loads require a bonded and unbounded specimens, the higher load rate detailed understanding of blast phenomena and the dynamic response of various structural elements. Luccioni, Ambrosini induced by impact resulted in the breaking of the and Danesi [5], performed a detailed analysis of the prestressed tendon. Compared to prestressed concrete, structural failure of a reinforced concrete building caused by reinforced concrete has received much more attention 190 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 a blast load. However, most of the current numerical magnitudes of the internal forces and deflections and modeling research is involved with massive computational rotation of the central line of beam or plate sections of the time and the model is difficult to build due to its complexity. structure. Evaluating a structure on the basis of the Therefore, for designers, it is imperative to establish a simple rotation of its sections is a methodology under modeling method to study the detailed behavior of the development at present, and is in accordance with recent building after the blast denotation. Techniques developed by research trends. The authors have used limit rotation Feng Fu [6], with ABAQUS, a 3-D finite element model representing a 20-storey building was built to perform the values (failure angle) determined experimentally on the blast analysis, a simplified direct simulation method of blast basis of the explosion load of masonry, reinforced load is applied, the nonlinear material behavior and dynamic concrete and window glass plates, comparing their own effects are also included in the simulation. results with results published by other authors. The collapse of a tall building under blast loading can also affect the adjacent buildings. Alex et al. [7], provided an 2. Blast Loading accurate prediction of the effects of adjacent structures on the blast loads on a building in urban terrain. In the paper, a An explosion is a rapid release of stored energy tentative attempt has been made to characterize the blast characterized by a bright flash and an audible blast. Part of environment by considering a simple urban configuration the energy is released as thermal radiation (flash); and part is with a relatively long, straight street segment and a T- coupled into the air as air blast and into the soil (ground) as junction at the far end. Numerical simulations using a ground shock, both as radials expanding shock waves. The computational Fluid dynamics (CFD) code Air3D has been rapid expansion of hot gases resulting from the detonation of used to determine the blast effects on building in a typical an explosive charge gives rise to a compression wave called urban terrain. Each simulation provided the variation with a shock wave, which propagates through the air. The blast distance of peak overpressure and impulse. When compared wave instantaneously increases to a value of pressure above with the corresponding variations for a surface burst of a the ambient atmospheric pressure. This is referred to as the hemispherical charge in a free-field environment, these positive phase that decays as the shock wave expands variations in the calculation of the pressure and impulse outward from the explosion source. After a short time, the enhancement factors at each scaled distance from the charge. pressure behind the front may drop below the ambient The resulting enhancement factors effectively modify the pressure, Fig.1. During such a negative phase, a partial blast parameters obtained from simplified analytical vacuum is created and air is sucked in. This is also techniques. accompanied by high suction winds that carry the debris for long distances away from the explosion source. Blast load parameters are reasonably easily determined for rectangular columns and can be derived from either the literature or numerous utility programs. Qasrawi et al. [8], investigated the pressure distribution around a circular column constructed in AUTODYN a numerical model. The model was verified and showed good agreement with established values. It was fund that the column radius increased the maximum reflected pressure at the point closest to the blast approached a maximum of approximately 0.9 of the design value quickly. It was also found that the pressure varied sinusoidal from this maximum to a minimum at the Fig. 1. Blast wave propagation (Abaqus 6.10) side of the column approximately equal to the incident pressure. The used a sinusoidal function to fit the As the shock wave travels outward from the charge, the distribution around the column with good results and this pressure in the front of the wave, called the peak pressure, steadily decreases as shown in Fig. 2. At great distances from curve fit can be used to find an equivalent design value. A the charge, the peak pressure is infinitesimal, and the wave 3-D dynamic analysis of an entire structure is used to can be treated as a sound wave. The observed characteristics determine the effects of an explosion and the response of of air blast waves are found to be affected by the physical the structure. D. Makavicka [9], discussed about the properties of the explosion source. At the arrival time tA, methodology for dynamic response assessment and its following the explosion, pressure at that position suddenly application to the new RC building. The authors used a increases to a peak value of overpressure, Pso, over the specific building as an example to illustrate the problem of ambient pressure, Po. The pressure then decays to ambient an explosion and the threat to the safety of the structure level at time td, then decays further to an under pressure P - so due to the explosion of a rather large terrorist charge (creating a partial vacuum) before eventually returning to installed in a car and initiated on a road adjacent to the ambient conditions at time td+ t - d . The quantity Pso is usually building. The structure response was assessed on the basis referred to as the peak side-on overpressure, incident peak overpressure or merely peak overpressure, TM 5-1300 [10]. of the results of 3D dynamic calculation using the 191 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 1772 114 108 P    (4) SO 3 2 Z Z Z In this paper the author used Mills formula shown in the Eq. 4, to calculate the peak overpressure. So the value of pressure calculated for 1m and 2m of stand points are respectively PSO = 1.766MPa and PSO = 0.25MPa. 3. MODELLING 3.1. Numerical Analyze Fig. 2. Blast wave pressure – Time history (Abaqus 6.10) The structure selected for this study is a 6-storey reinforced concrete building at the height of 18.5m. The 2.1. Blast Wave Scaling Laws length arranged as 6m in the longitudinal direction, and 6m span in the transverse direction. The storey height is 3.5 m All blast parameters are primarily dependent on the for the first storey and a standard storey beyond the first amount of energy released by a detonation in the form of a storey is 3.0m. The plan view and structural configuration of blast wave and the distance from the explosion. A universal the building are shown in Fig.3.The structural system of the normalized description of the blast effects can be given by building was made up of R.C Frames and Infill frames. The scaling distance relative to (E/Po)1/3 and scaling pressure typical beam size used is 350x250mm; the size of the relative to Po, where E is the energy release (kJ) and Po the columns is 300 x 300 mm. The rebar used for the concrete ambient pressure (typically 100 kN/m2). For convenience, beam section are 2ϕ16 and 2ϕ14 and for the column are however, it is general practice to express the basic explosive 4ϕ16.Columns are spaced at 5 m in X direction and 6m in Y input or charge weight W, as an equivalent mass of TNT. direction and are connected with beams and thickness of the Results are then given as a function of the dimensional 1/3 slabs is 175mm. Computer Modeling of the building was distance parameter (scaled distance) Z = R/W , where R is performed using the finite element software ABAQUS. The the actual effective distance from the explosion. W is 6 storey reinforced concrete building were frame structure of generally expressed in kilograms. Scaling laws provide columns, beams, and slabs. The columns and beams were parametric correlations between a particular explosion and a modeled as frame elements while the slabs were modeled as standard charge of the same substance. plate elements. The building model was assigned fixed bottom Support condition while a rigid diaphragm constraint 2.2. Prediction of Blast Pressure was allotted to all floors. Blast wave parameters for conventional high explosive materials have been the focus of a number of studies during the 1950’s and 1960’s. Estimations of peak overpressure due to spherical blast based on scaled distance Z = R/W1/3 was introduced by Brode (1955) as: 6.7 P  1barP  (1) S0 3 S0 Z View plan 0.975 1.455 5.85 PSO     0.019 bar (2) Z 2 3Z Z (0.1  P SO  10bar) Newmark and Hansen (1961) introduced a relationship to calculate the maximum blast overpressure, Pso, in bars, for a high explosive charge detonates at the ground surface as: W W 1/2 P (3) SO  6784  93( )3 3 R R Front and side elevation Fig. 3. 6-storey building model Another expression of the peak overpressure in kPa is introduced by Mills (1987), in which W is expressed as the 3.2. Analytical model equivalent charge weight in kilograms of TNT and Z is the scaled distance, as follow: In this paper, finite element analysis is performed using 192 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 the general purpose finite element package ABAQUS Where, I1 = σ1 + σ2 + σ3, c= cohesion, ϕ = internal /Explicit version 6.10 ABAQUS /Explicit solves dynamic friction angle. response problems using an explicit direct-integration With the purpose of obtaining the classical procedure. In an implicit dynamic analysis, the integration representation, this yielding surface can be expressed in the operator matrix must be inverted and a set of nonlinear following way: equilibrium equations must be solved at each time increment. In an explicit dynamic analysis displacements (1  ) 2 2  (1  ) 2  (  )2  A2(I  B)2  0 (6) 3 2 3 1 are calculated in terms of quantities that are known at the beginning of an increment; therefore, the global mass and Where: stiffness matrices need not be formed, which means that each increment is relatively inexpensive compared to the 2 2sen 3ccos A  ’ B  and compression increments in an implicit integration (Simulia, [11]). 3 sen sen Therefore, explicit method is more efficient than the stresses are considered positives in the main stress space. implicit integration method for solving extremely short- term events such as blast, explosion and impact. Fig.5 represents the classical quadric surface for a Solid elements in 3D were used to model the deformable conventional concrete (fc=25 MPa) with c =4.47 MPa, φ = structure presented in this paper. In this case, C3D8R: 8- 30º. The axis of the cone takes the direction of the vector (1, node linear brick, reduced integration, hourglass control, 1, 1). 3D Stress explicit was used to model frame Where the influence of mesh size has been studied and is sufficiently fine to ensure the accuracy of Models. The general approach for solving the non-linear modal analysis, for most realistic results a very small time step is required to obtain a stable solution. In this paper a time step of 0.05ms has been used to make the simulation. The building is designed to resist lateral loads due to wind and seismic ground motion specified by Turkish Regulations on Building. The Fig.4 shows the attack phenome at the Fig. 5. Graphic representation of the DP criterion structural configuration of the blast attack. The properties used for the concrete are as shown in the Table 1. Table 1. Reinforced concrete mechanical properties Density Elastic Poisson’s Compression Tensio (kg/m3) modulus ratio Strength n (MPa) (MPa) Strengt Fig.4. Structural configuration h (MPa) 4. Material Property 2400 2.6000 0.20 25 4 a) Concrete b) Steel There are several models available for modelling the adhesive based on different criteria for plastic deformation of The properties used for the steel are as shown in the concrete. Predictions of joint performance at large strains Table 2. close to joint failure depend on the model used. For the Table 2. Reinforcement steel mechanical properties prediction of failure, stress and strain distributions in the adhesive need to be accurately calculated and a failure Elastic Plastic criterion for the adhesive needs to be established. Density Young’ Yield Plastic In this studies Drucker Prager has been used which is (kg/m3) s Poisson’s Stress Strain defining as follow: modulu ratio (MPa) (MPa) s (MPa) The Drucker–Prager yield criterion (DP) is a pressure- dependent model for determining whether a material has 7.85E- 210.000 0.30 350 0 failed or undergone plastic yielding. The yielding surface of 009 the DP criterion may be considered depending on the internal friction angle of the material and its cohesion. In the space defined by the principal stresses, it is expressed as: 2.sen  2 2 6.c.cos (5)  .I1  1/ 2 ( 1  2 )  ( 1  3 )  ( 2  ) 2   0  3 3 sen  (3 sen) 193 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 5. Results and discussions In this study the attack has been occurred outside of the structure so the type of the explosion is classified as the open air explosion, Tehnički vjesnik (2012), which causes a wave that spreads from the source of detonation to the structure without any wave amplification. Due to the blast source’s distance and height is away from the structure, the explosion provoked a wave increase due to the reflection of the ground before it contacts the structure. Also the explosion occurred Fig. 7. Nodal displacement curve on the ground and the initial pressure is immediately After 0.125 ms the strain energy reached the maximum increased as a result of refraction on the ground. value of 0.146 MPa, in the case: Pos=1.766MPa (standoff Due to the evolving of high nonlinear behavior and the distance of 1m) and got decrease till the failure point as complexity of the model, to analyses the effect of blast shown in the Fig.8 in color blue. Also the same observation loading, the column located on the left side of the attack at has been made on second curve which also increased from the ground floor has been selected. The Fig.6 illustrated the 0.006 MPa at 0.025ms and remain virtually constant. blast loading after the explosion and selected column deflection. Fig. 8. Strain energy The stress reach the peak value of 28, 9617 MPa at 0,01ms in the case: Pos=1.766MPa (standoff distance of 1m) and only at 0.015 reach the peak value of 25.7434 MPa in the second case as shown in the Fig. 9. a) 3D model after the attack Fig. 9. Stress The energy release after the explosion for the whole model reached the maximum value of 0. 4 E10 J, after 0.0085 ms as shown in the Fig.10, from zero and at the same moment started decaying till the end of the simulation. b) A column displacement expressed in mm Fig. 6. 6- Storey building modelled in ABAQUS The analysis has done according to the peak pressure and standpoint calculated above. For each peak pressure and stand point the model has been simulated. It can be seen in the Fig.7 that there is a significant difference between two Fig. 10. Total energy. graphics in which the maximum value of displacement in the 6. Conclusion nodal shown above in the Fig. 6 (b) is: 0.136m for the case POS =1.766MPa (standoff distance of 1m) and 0.037 for the According to the results the system affects significantly case POS=0.25Mpa (2m). when the peak pressure (which is proportional to the charge 194 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Namata S. Sabiou et al., Vol.3, No.4, 2017 weight) increases and standoff distance decreases [3] J. Magnusson & M. Hallgren (2004). “Reinforced High respectively. Strength Concrete Beams Subjected to Air Blast Loading”, WIT Transactions on The Built So the standoff distance is the key parameters that Environment, ISSN: 1743-3509(73), 10. determines the blast pressure which decrease so far the effect of blast loading on the structure can be reduce by keeping or [4] Luccioni, B.M. and Luege, M. (2006). “Concrete prevent the bomb attack as far away as possible by Pavement Slab under Blast Loads”, International Journal maximizing the standoff distance. of Impact Engineering 32(8): 1248-1266. Blast has a characteristic of high amplitude, the results shows [5] B.M. Luccioni, R.D. Ambrosini and R.D. Danesi, (2004). that columns subjected to high pressure they could cause big “Analysis of Building Collapse under Blast Loads”, deformation and exceed the support reaction so the columns Engineering Structures 26 (2004) 63–71. which are close to explosion are damaged and can be lost. [6] Feng Fu, (2009). “Advanced Modeling Techniques in Ones the critical load bearing columns lost which leads to Structural Design”, Wiley Blackwell, City University sudden changes in the building geometry and load path London. which initiates a chain reaction of structural elements failure [7] Alex M. Remennikov and Timothy A. Rose (2005), The value of the strain energy is higher in the case of major “Modelling blast loads on buildings in Complex city peak pressure because of the higher lateral deflection geometries”, Computers & Structures, 83 (27), 2197- presented, which shows better energy absorption. 2205. References [8] Y. Qasrawi, P. J. Heffernan & A. Fam (2010). “Numerical Determination of Reflected Blast Pressure [1] N. Ishikawa, H. Enrin, S. Katsuki and T. Ohta (1998), Distribution on Round Columns”, WIT Transactions on “Dynamic Behaviour of Prestressed Concrete Beams the Built Environment, Vol 113, SN 1743-3509. under Rapid Speed Loading”. Department of Civil [9] Daniel Makovicka, Daniel Makovicka, Jr. (2010). Engineering, National Defence Academy, 1-10-20 “Simplified Evaluation of a Building Impacted by a Hashirimizu Yokosuka 239, Japan. terrorist Explosion”, WIT Transactions on The Built [2] N Ishikawa,S Katsuki &K Takemoto(2002), “Incremental Environment, vol. 113, 93-103. Impact Test And Simulation Of Prestressed Concrete [10] TM 5-1300 Structures to Resist the Effects of Beam”, Structures Under Shock and Impact VII, N Jones, Accidental Explosions, US Army, November 1, 1990. CA Brebbia and AM Rajendran , ISBN 1-85312-911- 9(63),10. [11] ABAQUS User’s Manual, Version 6.10, Simulia 2010. 195 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017 The Investigation of Wear Behaviour of Dual Phase Steel DP600 Yasin Aygül*, İsmail Ovalı**‡ *Department of Manufacturing Engineering, Faculty of Technology, Affiliation of Pamukkale University, yasin_aygul@yahoo.com ** Department of Manufacturing Engineering, Faculty of Technology, Affiliation of Pamukkale University, iovali@pau.edu.tr (yasin_aygul@yahoo.com, iovali@pau.edu.tr) ‡ İsmail Ovalı; Yasin Aygül, iovali@pau.edu.tr, yasin_aygul@yahoo.com, Tel: +90 258 296 4138, Fax: +90 258 296 41 96, iovali@pau.edu.tr Received: 23.07.2017 Accepted: 13.12.2017 Abstract- Dual phase steels have been widely used in many applications in the automotive industry. In this study, the wear behaviour of commercially available DP600 steel with dual-phase structure are investigated. The wear rate, volume loss and friction coefficient are used to optimize wear resistance of DP600 steel with dual-phase. Abrasive and Adhesive wear tests were carried out to determine the abrasive friction coefficient of the DP600 steel. Experimental studies show that the ferrite and martensite phases based on microstructures of the DP600 steel with dual-phase structure have an effect on tensile and wear tests. The wear resistance of DP600 steel can be controlled and optimized with wear condition (the wear rate, volume loss and friction coefficient). Keywords Dual Phase Steel, Rolling, Flat Product, Wear, Abrasive, Adhesive. 1. Introduction composition of test samples; 0,065% C, 1-1,2% Mn, 0.9-1% Mo, and 0.05% Al. As a result of the experiment; Samples With the aim of meeting the needs and expectations of containing martensite in high amounts showed dual-phase the industrial area, many new researches are made by the steel properties. Also, contrary to what is expected, they material engineers day by day in the steel industry. As a showed double-phase steel properties in samples having 70- result of these studies, steel materials with a light, suitable 90% polygonal ferrite, 0-25% bainitic ferrite and martensite strength, good deformation ability and high forming ability volume above 10% of the steel they produced. The study are foregrounded [1,2]. shows that microstructures having the above-mentioned characteristics can be produced when appropriate cooling Dual-phase steels are produced by the critical annealing- conditions are met. As a result of the experiment, there is no quenching process of High Strength Low Alloy (HSLA) significant relationship between residual austenite and steels. It is a low-alloy or non-alloy high-strength steel type mechanical properties [2,3]. containing martensite phase dispersed in ferrite in microstructure. Microstructures also contain trace amounts of Hansen and Bramfitt investigated two different experiments to austenite, bainite and pearlite. Basic dual phases; Ferrite produce dual phase steel. Critical annealing heat treatment and hot phase and martensite phase. These steels show many ideal rolling. The chemical composition of the test sample; 0,04-0,11C, 0,80-1,2Mn, 0,51-1,45Si, 0,05-0,65Cr, 0,27-0,62Mo. The Dual- properties when compared to HSLA steels. These; High phase steels produced shows similar tensile properties except yield strength, homogenous forming, percent elongation properties strengths. The yield strength of hot rolled steels is higherThis [1,2]. is caused by the presence of spherical bainite, which forms Coldren and Cornford have experimented to produce the third phase, except polygonal ferrite and martensite, double-phase steel during ingot hot rolling. Chemical which are found in the microstructure. They stated that the 196 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017 coiling temperature should be below 610 ° C in order to rolling method, the formation of the structure is difficult to produce martensite without pearlite formation [2,4]. follow. CCT diagrams are used for this [2,3]. In this study; Abrasive and adhesive wear behaviours of The recommended alloying elements are Mn, Cr, Ni dual phase DP600 steel produced for industrial purposes and Mo. If quenching is carried out after hot rolling, double- were investigated. In addition, the yield stress, tensile stress phase steel structure is obtained with less alloy addition and elongation properties of selected specimens of different [2,3]. orientations of DP600 Steel (Orientation) were examined. 2.3. Tensile Test 2. Experimental Studies To make the tensile test, the drawing sample is prepared 2.1. Material according to the standards from the material. In this study, test specimens were prepared according to TS 138 EN In the experimental work, DP600 dual-phase steel with a 100008 standards and experiments were carried out. The thickness of 4.5 mm produced by Hot Rolling was selected. selection direction of samples is given in Fig. 2.2. Sheet The chemical composition of steel is given in Table 1. metal materials exhibit different mechanical properties in different directions (isotropic and anisotropic) The samples Table 1. Chemical composition of the sample (%Wt) are angularly removed from the plate. (R0, R45, R90) Angular Chemical composition (%Weight) samples are prepared and necessary calculations are made [6]. %C %Mn %Si %P %S %Al %V %Ti 0,079 0,82 0,23 0,039 0,0013 0,031 0,001 0,002 %Nb %Mo %Ca %N %Cu %Ni %Cr %Ti/N 0,000 0,002 0,0011 0,0067 0,02 0,04 0,66 0,3 2.2. Rolling Production of DP600 steel with dual phase structure used as test samples were carried out by Hot Rolling method. The production line is given in Fig.2.1. Dual-phase steel Fig. 2. Tensile test sample selection production in the Hot Rolling process is more economical than other production methods [2,3]. 2.4. Wear Test Load on the sample is recorded during wear test in order to get data for calculation of friction coefficient. Eq. (1) was used for defining friction coefficient. F Fig. 1. Hot Rolling process   The coefficient of friction is,     (1)  P  Production of flat rolled products by hot rolling is carried out as a result of applications at different stations in Where F is the frictional force measured by the load cell Figure 2.1. It is annealed to hot shaping temperature in slab. and P is the normal load on the specimen [7]. After the annealing process, the slab enters the reaction with oxygen. The oxide layer formed in this reaction is called The volume loss was calculated from the weight loss descaling. Pressurized water is sprayed onto the slab to break according to the following Eq. (2). the resulting oxide layer. The slab, which is cleaned from the oxide layer on the surface, is first sent to the coarse rolling process and then to the coil box to prevent temperature loss.     The slab is subjected to strip rolling according to the order 3  WeightLossg   sizes. The slab is then cooled down continuously in the Volume Loss (m m )= 1000 (2)   shower tables controlled by the automation system to create  g  Density 3   the desired microstructure. The strip is then coiler at the   mm   appropriate temperature [5]. The wear tests were conducted five times for every sample. The rate of cooling of material during rolling is too The wear rate was calculated from the following Eq. (3) [7]: slow to bring ordinary carbon steels to normal dual-phase steel. At slow cooling rates after rolling, suitable alloying elements should be used to form an ideal dual-phase steel structure. In the production of dual-phase steel by the hot 197 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017  VolumeLossmm3   2.4.2. Adhesive Wear Test Wear Rate (mm3/m)=  1000 (3)  SlidingDistancem   Load Experiments have been carried out on a pin on disk abrasive wear device designed according to ASTM G99- Abrasive and adhesive wear tests were carried out on the 05 as standard. The wear tester used in the wear tests is given TURKYUS Pin-On-Disc. The weights and densities during in Figure 5. The surfaces to be abraded of all samples were the test were measured on the RADWAG electronic scale sanded with 220 mesh SiC abrasive before testing to obtain [7]. the same surface quality. The wear tests were carried out at a speed of 3.16 m / s with 5, 10 and 15 N loads. In the experiment; AISI 52100 Bearing Steel material is used as abrasive. AISI 52100 bearing steel has 60 HRC hardness. Experimental times were determined as 95,190 and 285 meters. The specimens were moved perpendicular to the abrasive disc surface. The weight loss of the samples was also measured with an electronic balance of ± 0.001 g precision. Furthermore, in order to determine the friction coefficient, the frictional forces were measured with the load sensor mounted on the device and the friction coefficients Fig. 3. Pin-On-Disk Wear Device were determined. Before and after the wear test, the samples were cleaned with ethyl alcohol and dried. All samples were 2.4.1. Abrasive Wear Test subjected to wear test twice as shown in Table 3 and weight loss, volume losses, friction coefficients and wear rates were Experiments have been carried out on a pin on disk calculated using the formulas (1,2,3) indicated by the abrasive wear device designed according to ASTM G99-05 averages of these values [7]. as standard. The wear tester used in the wear tests is given in Figure 5. The surfaces to be abraded of all samples were Table 3. Adhesive test parameters. sanded with 220 mesh SiC abrasive before testing to obtain the same surface quality. The wear tests were carried out at a Sample Sliding Applied Load Wear Distance Numbers Distance (N) (m) speed of 3.16 m / s with 5, 10 and 15 N loads. 220 mesh SiC (m/s) was used as the abrasive disc. Experimental times were determined as 95,190 and 285 meters. The specimens were DP-600 3,16 30 475 moved perpendicular to the abrasive disc surface. The weight DP-600 3,16 30 575 loss of the samples was also measured with an electronic DP-600 3,16 30 665 balance of ± 0.001 g precision. Furthermore, in order to DP-600 3,16 40 475 determine the friction coefficient, the frictional forces were DP-600 3,16 40 575 measured with the load sensor mounted on the device and the DP-600 3,16 40 665 friction coefficients were determined. Before and after the wear test, the samples were cleaned with ethyl alcohol and DP-600 3,16 55 475 dried. All samples were subjected to wear test twice as DP-600 3,16 55 575 shown in Table 2 and weight loss, volume losses, friction DP-600 3,16 55 665 coefficients and wear rates were calculated using the formulas (1,2,3) indicated by the averages of these values 3. Experimental Results [7]. 3.1. Tensile Test Results Table 2. Abrasive test parameters Sample Sliding Applied Load Sliding The results of three different tensile tests of the two- Numbers Distance (N) Distance (m) phase DP600 steel specimen taken in three different (m/s) directions are given in Table 4. DP-600 3,16 5 95 DP-600 3,16 5 190 Table 4. Tensile Test Results DP-600 3,16 5 285 Sample DP600 DP600 DP600 DP-600 3,16 10 95 R0 R45 R90 DP-600 3,16 10 190 DP-600 3,16 10 285 Thickness (mm) 4,19 4,16 4,20 DP-600 3,16 15 95 Width (mm) 25,37 25,13 25,19 DP-600 3,16 15 190 Yield Strength (MPa) 347 359 341 DP-600 3,16 15 285 Tensile Strength (MPa) 579 575 572 Elongation (%) % 34 % 31 % 32 198 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017 Tensile test results show that there are very similarities in loss. 40N, 55N test specimens show similarities in volume the yield, tensile break results of three differently prepared losses throughout the test period. However, it can be seen (R0, R45, R90) samples. The microstructure formed in the that the weight loss profile of 30N test sample is less than the sample produced by hot rolling shows the same behaviour in other samples. Our test sample of 55N and 40N showed a all the sections as isotropic regardless of direction. noticeable difference from the other sample at 500m, since the force applied by the abrasive was higher. Adhesive wear Tensile testing of flat steel products produced by hot tests have less volume losses than abrasive wear tests due to rolling; It is related to deformation rates, annealing, supply abrasive . and winding temperatures [8]. The most important parameters determining the yield and tensile strengths in the double-phase steels are the proportions of the ferrite and martensite phases present in the microstructures. The equivalent strength of low-alloy steels, which are used equivalently for dual-phase steels, exhibit approximately the same behaviour as the yield and draw characteristics. 3.2. Wear Test Results To characterize the wearing behavior of the DP600 Steel; Volume loss, wear rate and coefficient of friction were evaluated. 3.2.1. Volume Loss of Abrasive Sample Fig. 5. Volume Loss of Adhesive samples The volume of adhesive sample in Fig. 6 show good correlation within the literature [9]. 3.2.3. Wear Rate of Abrasive Sample Fig. 4. Volume Loss of Abrasive samples Volume loss of the abrasive wear test of DP600 Steel under different loads (5N, 10N, 15N) are given in Fig.3.1. The volume loss resulting from the increase in load also increased significantly in the Fig. 6. It can be seen clearly that abrasive wear tests have higher volume loss rates than adhesive wear tests due to abrasives. 5N, 10N and 15N test Fig. 6. Wear rate of Abrasive Samples specimens have similar volume losses up to 200 m. In the range of 200-300 m there is no similarity in volume losses of In the characterization of wear behaviour of steel, the the samples. The result shows the good correlation of volume losses are not enough alone. In order to determine literature study [9]. the effect of sliding distance on wear behaviour, wear rate data should be evaluated. 3.2.2. Volume Loss of Adhesive Sample Fig 6 exhibits abrasive wear test wear rates of the DP600 In The Fig.3.2 shows the volume loss adhesive wear test Steel under different loads (5N, 10N, 15N). 10N and 15N of DP600 Steel under different loads (30N, 40N, 55N) Figure samples showed a similar wear rate at the 300 m sliding 7 shows that volume loss during the test increases distance. In other words, the high sliding distance does not significantly at the end of the load increase Similarly, the have a significant effect on the 10N and 15N samples. 5N, increase in sliding distance has led to an increase in volume 10N and 15N samples showed higher wear rates than the 199 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017 adhesive wear test. This is due to the use of 220 mesh SiC Tests were carried out on the graph with different loads (5N, discs which are used as abrasive [7]. 10N, 15N). As a result of the abrasive wear test of the DP600 Steel, the friction coefficient was plotted depending on the 3.2.4. Wear Rate of Adhesive Sample sliding distance. In Figure 9 it has been observed that the friction coefficient decreases considerably with increasing force. This result is parallel to the studies in the literature. The highest coefficient of friction was obtained at ~ 0.8 to 5N. The lowest coefficient is seen in the ~ 0.4 to 15N sample. 3.2.6. The Friction Coefficient of Adhesive Sample Fig. 7. Wear rate of Adhesive Samples Wear test were carried out on the graph with different loads (30N, 40N, 55N). Fig. 8 illustrates the of wear rates of adhesive samples. It can be seen that the wear rates obtained as a result of the adhesive wear tests are lower than the abrasive wear rates. This result; Is an abrasive disk type used Fig. 9. Friction Coefficient of Adhesive Samples in adhesive wear tests. 40N and 55N samples exhibit similar wear rates between 300-500 m sliding distances. 35N test The friction coefficient of adhesive samples is given in sample showed a proportional increase between 300-500 m. Fig. 9. It can be clearly understood that the friction Showed a remarkable increase in the range of 500-800 m. coefficient is affected by sliding distance. Although the loads on the adhesive wear test are higher than those on the 3.2.5. The Friction Coefficient of Abrasive Sample abrasive wear test, there is a significant reduction in their resistance. This is because; The disc used as an abrasive has a metallic structure. All of the samples are beginning to test with a significant increase before 50 mm sliding distance. A decrease in the coefficient of friction is seen with increasing force. 4. Conclusions In this study; The mechanical properties of commercially available dual phase DP600 steel; Tensile and yield strengths, abrasive, adhesive wear behaviors were investigated. The results obtained are given below.  Tensile tests were carried out by extracting tensile samples in different rolling directions of the flat test specimen produced after rolling. Tensile strength, tensile Fig. 8. Friction Coefficient of Abrasive Samples strength and breaking elongation of the specimens selected depending on the direction of orientation were The most important parameter used in studying the wear determined at the end of the tensile test. behaviour of metals is the friction coefficient. The coefficient  All samples (isotropic) showed the same behavior of friction is used to characterize how the metals exhibit regardless of the direction of the sample used as a result resistance during wear. of tensile tests.  As a result of the wear test, volume loss, wear rate and It can be related the coefficient of friction to the force of friction coefficients were determined under the slip breaking a piece from a material. If it is easy to break off the distance and different loads. In the results of wear material surface, the friction coefficient decreases. If the obtained; The volume losses and wear rates found in the applied force increases, it should be easier to break the piece. 200 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Yasin Aygül et al., Vol.3, No.4, 2017 abrasive wear differ from the results of the adhesive wear. [4] Hansen S.S. & Bramfitt B.L., “Hot Strip Mill Processing  The highest abrasive rates as a result of abrasive wear of Dual-Phase Steels” int Conf. On steel Rolling, Sci, and tests; High sliding distance distance and 15 N load. Tech. Of flat Roolled Poducts, ISIJ, vol:4, no:7, pp:1297-  The highest Adhesive rates as a result of abrasive wear 1308, Tokyo, 1980. tests; High sliding distance distance and 55 N load. [5] İSDEMİR Sıcak Haddeleme Tesisleri, Hatay (2016) References [6] Gavas, M., Aydın, M., Yaşar, K., Altunpak, Y., “Üretim Yöntemleri ve İmalat Teknolojileri”, Ankara, 2013. [1] Rashid M.S., “GM980X – A Unique High Strenght Sheet Steel with Superior Formability”, SAE, Preprint 760206, [7] Ovalı, İ., “Küresel Grafitli Dökme Demirlerin Yüzeyinde (1976). Çil Oluşumu ve Östemperleme Isıl İşleminin Mikroyapı ve Mekanik Özellikler Üzerine Etkisi”, Gazi Üniversitesi Fen [2] Demir, B., “Ereğli Demir ve Çelik Fabrikaları Sürekli Bilimleri Enstitüsü Doktora Tezi, Ankara, 2012. Tavlama Hatlarında Çift-Fazlı Çelik Üretilebilirliğinin İncelenmesi”, Doktora Tezi, Gazi Üniversitesi, Fen Bilimleri [8] Çapan, L., “ Metallere Plastik Şekil Verme ” Çağlayan Enstitüsü, 159 s., Ankara, 2003. Kitapevi İstanbul, 2003. [3] Coldren A.P. & Eldis T., “Using CCT Diagrams to [9] Andres, M,R,. “Tribological behaviour of DP600 dual Optimize the Compositions of an as-Rolled Dual Phase phase steel on uni- and bi-directional discontinuous sliding Steels”, Jornal of Metals, vol:4, no:4, pp:41-48, 1980. wear”, Madrid, 2015. 201 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Karakaya and Ucun et al., Vol.3, No.4, 2017 Binding Energy and Stability Calculations on Hydrogenated Forms of Substituted Carbazoles as Hydrogen Storage Materials Mustafa Karakaya*‡, Fatih Ucun** *Department of Energy Systems, Faculty of Engineering & Architecture, Sinop University, Sinop 57000, Turkey **Department of Physics, Faculty of Arts and Sciences, Süleyman Demirel University, 32260 Isparta, Turkey (mkarakayafizik@hotmail.com, fatihucun@sdu.edu.tr) ‡ Corresponding Author; Mustafa Karakaya, Sinop University, Sinop 57000, Turkey, Tel.: +90 368 2715516 Fax: +90 368 2714152, mkarakayafizik@hotmail.com Received: 15.08.2017 Accepted: 13.12.2017 Abstract- The aim of this work is to explore the stable hydrogenated forms of carbazole and 9-methylcarbazole molecules by using M06-2X density functional as computational method. Binding energies per hydrogen atom in these hydrogenated forms were calculated by the counterpoise correction procedure. Relative energies, complexation and binding energies for the conformers of dodecahydrocarbazole were also calculated. Stabilities of all the hydrogenated forms were discussed by the analysis of the frontier molecular orbitals. Keywords: Hydrogen energy technologies; hydrogenation; M06-2X theory; binding energy. 1. Introduction respectively, for the organic liquid hydrides [2, 9‒11]. Hydrogenation is the process of adding hydrogen under Hydrogen is an important energy source that is shield to pressure with the aid of a catalyst. The aim of this procedure the global climate changes against the ever-increasing energy is that the hydrogen atoms attend to multiple bonds. In the demands [1]. Hydrogen technologies relates to development hydrogenation reaction, alkane is converted to alkane of new methods in areas such as obtain the hydrogen, storage condition by addition of the hydrogen to the unsaturated and processing. Also these technologies give very effective carbon chemical bond. The dehydrogenation process is the results in science to prevent negative environmental impacts opposite of this. Hydrogenation and dehydrogenation and create green energy sources. Hydrogen storage methods kinetics and catalysis of carbazole derivatives and influences are classified as pressurized hydrogen, carbonaceous of the reaction temperature and pressure have been materials, metal and complexation hydrides, organic liquids experimentally studied in the literature [12‒14]. according to storage capacity (wt%), safety, stability and In this study we have theoretically aimed to investigate transaction costs [2‒6]. and interpreted the hydrogen binding energies, complexation Carbazole structures in liquid organic hydrides are energies and energy gaps between frontier (highest occupied noteworthy with reversible hydrogen sorption characteristics and lowest unoccupied) molecular orbitals on the as hydrogen storage materials [7]. Liquid organic hydrides hydrogenated forms of substituted carbazoles as hydrogen have significant advantages in hydrogen storage as access to storage materials via computational methods. Before high gravimetric hydrogen storage level, lack of carbon anything else, we have determined the optimized molecular monoxide and carbon dioxide emissions as by-products [7, forms of carbazole (S1), 9-methylcarbazole (S2) and 8]. Hydrogenation and dehydrogenation techniques are used octahydro-, dodecahydro- structures as their hydrogenated in the process of hydrogen storage and releasing, forms. The next phase of our study consists of the 202 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Karakaya and Ucun et al., Vol.3, No.4, 2017 evaluations about the relationship between hydrogen complexation and binding energies were calculated by the binding energies, frontier molecular orbitals and hardness of counterpoise correction procedure (CP) [17]. all the S1-S2 hydrogenated forms to determine their chemical 3. Results and discussion reactivity and stability. Fig.1 is inclusive the optimized molecular structures of 2. Computational details S1, S2 and their hydrogenated forms obtained by the output All the quantum calculations on the structural optimizing of quantum computing. In the figure the hydrogens of which processes and molecular orbitals were performed by using the binding energies will be calculated, are located in the M06-2X theory combined with 6-31+G (d,p) basis set level. green zone. The minimum-energy structures of S1 and S2 Inputs on the atomic coordinates for all the geometry have -517.280 and -556.571 a.u. of energies at M06-2X optimizations were created by Gauss View program [15]. calculation method and 6-31+G (d,p) basis set, respectively. Optimizations and computations on the S1, S2 hydrogenated Calculated relative energies and hydrogen binding energies forms and the atomic hydrogen binding energies were which bind to carbons have been reported in Table 1 for the obtained by using Gaussian 09W software database [16]. substituted carbazoles and their hydrogenated forms. Basis set superposition error (BSSE) corrections, Fig. 1. Optimized structures of S1, S2 and their hydrogenated structures computed at M06-2X theory, 6- 31+G (d,p) level The average values of relative energies per hydrogen basic set. While the number of functions included in the atom are 16.24 and 16.28 eV for S1+nH2 and S2+nH2 forms, basic set increases, the BSSE energy is reduced. So, The respectively. Total binding energies for the hydrogen atoms BSSE energy has a small value in a large basis set. Another attached to carbons have been calculated as -45.815 and - factor that affects the BSSE energy value is occupied or 45.744 eV in the S1 and S2 structures, respectively. These unoccupied orbitals. Unoccupied orbital’s give greatly an e values are approximately -82.979 and -82.902 eV in the increasing effect in the BSSE energy. S1+6H2 and S2+6H2 hydrogenated forms, respectively. For the dodecahydrocarbazole structures in Fig. 1, we While the binding energies per hydrogen atom are -5.727 and like to examine the conformers for which the relative values -5.718 eV in the S1 and S2 structures, these results are - between the ground state energies are many minor by 4.149 and -4.145 eV in the S1+6H2 and S2+6H2, calculation method. The relative energies, complexation and respectively. binding energies for the conformers of On the other hand, the calculated BSSE energies are dodecahydrocarbazoles are given in Table 2. The positioning increasingly on the rise in hydrogenated forms of S1, S2. of the conformers has been identified at the bottom line of This situation can be attributed to positioning where there are Table 2 (see also Fig. 1). Considering the total energies, we involved fragments and of course, to the applied method and say that the most stable optimization is conformer 1 for the 203 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Karakaya and Ucun et al., Vol.3, No.4, 2017 both hydrogenated forms (+6H2). The calculated energy highest occupied and lowest unoccupied orbital energies and differences between the various conformations for +nH also molecules with high HOMO energies are more reactive hydrogenated forms of 9-ethylcarbazole have been calculated and of course unstable [22]. by Density Function Theory (DFT) method with the B3LYP Table 2. Calculated relative, complexation and hydrogen hybrid functional including electron correlation in a previous binding energy values for conformers of study [7]. dodecahydrocarbazole formed according to positions of Table 1. Calculated relative and hydrogen binding energy hydrogens attached to carbons. values for substituted carbazoles and, their hydrogenated Binding forms. Energy Relative Binding Energy** Complexation Relative Per Organic Energy Energy* Per Hydrogen Atom Hydrogen Liquid Energy [corrected] Atom Materials (eV) (eV) Conformers (eV) (eV) (eV) S1 0.000 -5.727 S1+6H2 +2H2 64.822 -4.732 1 0.000 -82.975 - 4 .149 +4H2 129.592 -4.242 2 0.2060 -84.312 - 4 .216 +6H2 195.785 -4.149 3 0.2062 -83.627 - 4 .182 4 0.3389 -84.219 - 4 .211 S2 0.000 -5.718 5 0.4403 -84.541 - 4 .227 +2H2 64.783 -7.080 6 0.5453 -83.828 - 4 .191 +4H2 130.559 -3.994 S2+6H2 +6H2 195.795 -4.145 1 0.000 -82.897 - 4 .145 * Relative values between the total energy 2 0.1690 -84.350 - 4 .218 ** obtained by using CP corrected energy. 3 0.2039 -83.862 - 4.193 As seen in Table 2 the minimal corrected complexation energies have been computed as -84.541 (conformer 5) and - 4 0.3860 -84.090 - 4.205 84.350 eV (conformer 2), approximately. The lower value of 5 0.5478 -84.252 -4.213 the complexation energy corresponds to the more stable complex [18]. Also, the conformer 5 in the S1+6H2 and the 6 0.4972 -83.773 -4.189 conformer 2 in S2+6H2 have minimal values as binding Conformer 1= b(front),acd(back); 2= abcd(front); 3=ab(front),cd(back); energy per hydrogen atom. In the both hydrogenation 4=abc(front),d(back); 5=ac(front),bd(back); 6=bc(front),ad(back) structures, the conformers 2 and 5 draw attention with their minimum complexation and binding energies. In here we take into account the conformers 1 for the dodecahydro- structures in the calculations and in the comments relating to binding energy and molecular orbitals since they have the minimum total energies. The first and second frontier molecular orbitals and energy diagrams of S2+6H2 hydrogenated form have been displayed in Fig. 2. The claret red and green tones represent the positive and negative wave functions, respectively, for the orbitals. The visual for the highest occupied molecular orbital is scattered over the nitrogen and the carbons close to the nitrogen. The positive wave functions for the lowest unoccupied molecular orbitals are focused in the C-H bonds. Frontier molecular orbitals (HOMO and LUMO), act as the electron donor and acceptor, respectively, and these orbital Fig. 2. Highest and lowest energy levels of occupied and energy values explain the optical and electrical properties of unoccupied orbitals for S2+6H2 hydrogenated forms the molecules in quantum computations [19-21]. As seen in Table 3, the HOMO energies have been HOMO-LUMO energy gaps and chemical hardness in computed at the highest values in the octahydro- structures. hydrogenated carbazoles are given in Table 3. The type of The energy gaps in FMO’s comment the intramolecular bonding interaction can be explained by the topic of the charge transfer interactions and, its lower value means a 204 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Karakaya and Ucun et al., Vol.3, No.4, 2017 more reactive molecule, chemically [23]. On the other hand, [2] Z. Jiang, Q. Pan, J. Xu, T. Fang, “Current situation and the higher values of hardness are directly related to the prospect of hydrogen storage technology with new stability [23-26]. The octahydro- structures have smaller organic liquid”, Int. J. Hydrogen Energ., vol. 39, pp. energy gaps relative to the others in this study. Conversely, 17442-17451, 2014. the maximal energy gaps have been computed in the S1+6H2 [3] P. Chena, M. Zhu, “Recent progress in hydrogen and S2+6H2 hydrogenated forms. The calculation results for storage”, Mater. Today, vol. 11, pp. 36-43, 2008. the chemical hardness support that the stabilities of S1+6H2 and S2+6H2 hydrogenated forms are highest. [4] S.G. Chalka, J.F. Miller, “Key challenges and recent progress in batteries, fuel cells, and hydrogen storage Table 3. HOMO energies, energy gaps and chemical for clean energy systems”, J. Power Sources, vol. 159 harnesses for substituted carbazoles and their (1), pp. 73-80, 2006. hydrogenated forms [5] W. Peschka, C. Carpetis, “Cryogenic hydrogen storage HOMO Chemical Organic and refueling for automobiles”, Int. J. Hydrogen energy Hardness Liquid Energy Gap Energ., vol. 5(6), pp. 619-625, 1980. Materials (eV) (eV) (eV) [6] E. Rönnebro, “Development of group II borohydrides as S1 -6.978 6.724 3.362 hydrogen storage materials”, Curr. Opin. Solid St. M., vol. 15(2), pp. 44-51 2011. +2H2 -6.616 6.850 3.425 [7] K.M. Eblagon, D. Rentsch, O. Friedrichs, A. Remhof, A. +4H2 -6.286 6.638 3.319 Zuettel, A.J. Ramirez-Cuesta, S.C. Tsang, +6H -7.565 7.902 3.951 “Hydrogenation of 9-ethylcarbazole as a prototype of a 2 liquid hydrogen carrier”, Int. J. Hydrogen Energ., vol. 35, pp. 11609-11621, 2010. S2 -6.811 6.571 3.286 [8] N. Kariya, A. Fukuoka, T. Utagawa, M. Sakuramoto, Y. Goto, M. Ichikawa, “Efficient hydrogen production +2H2 -6.480 6.773 3.386 using cyclohexane and decalin by pulse spray mode +4H2 -6.270 6.572 3.286 reactor with Pt catalyst”, Appl. Catal. A, vol. 247, pp. 247, 247-59, 2003. +6H2 -7.311 7.780 3.890 [9] R.H. Crabtree, “Hydrogen storage in liquid organic heterocycles”, Energy Environ. Sci., vol. 1,pp. 134-8, 2008. 4. Conclusion [10] F. Alhumaidan, D. Cresswell, A. Garforth, “Hydrogen storage in liquid organic hydride: producing hydrogen In present study, the hydrogenated forms of substituted catalytically from methylcyclohexane”, Energy Fuels, carbazoles have been achieved by the output of quantum vol. 25, pp. 4217-34, 2011. computing at M06-2X theory, 6-31+G (d,p) level. Considering the total energies, the most stable optimization [11] A. Shukla, S. Karmakar, R.B. Biniwale, “Hydrogen is conformer 1 for the S1+6H and S2+6H structures. But delivery through liquid organic hydrides: considerations 2 2 the conformers 2 and 5 have drawn our attention with their for a potential technology”, Int. J. Hydrogen Energ., minimum complexation (corrected) and binding energies per vol. 37, pp. 3719-26, 2012. hydrogen atom. As for the hydrogenated forms, S1+4H2 and [12] X. Ye, Y. An, G. Xu, “Kinetics of 9-ethylcarbazole S2+4H2 are more reactive than the other forms according to hydrogenation over Raney Ni catalyst for hydrogen the highest occupied molecular orbitals. Conversely, the storage”, J. Alloy. Compd., vol. 509, pp. 152–156, stabilities of the S1+6H2 and S2+6H2 hydrogenated forms 2011. are highest since of their maximal energy gaps. [13] F. Sotoodeh, B.J.M. Huber, K.J. Smith, Acknowledgements “Dehydrogenation kinetics and catalysis of organic heteroaromatics for hydrogen storage”, Int. J. This study was financially supported by Sinop University Hydrogen Energ., vol. 37, pp. 2715-2722, 2012. with the project number of MMF-1901.14-01. The author [14] F. Sotoodeh, B.J.M. Huber, K.J. Smith, “The effect of gratefully acknowledges the support of Sinop University, the the N atom on the dehydrogenation of heterocycles presidency of project management office. used for hydrogen storage”, Appl. Catal. A-Gen., vol. 419– 420, pp. 67– 72, 2012. References [15] R. Dennington, T. Keith, J. Millam, GaussView, [1] V. Balema, “Hydrogen Storage Materials”, Material Version 5.0.9, Semichem Inc., Shawnee Mission, KS, Matters, vol. 2, pp. 1-31, 2007. 2009. 205 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Karakaya and Ucun et al., Vol.3, No.4, 2017 [16] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, [22] H. Ullah, A.A. Shah, S. Bilal, K. Ayub, “DFT Study of M.A. Robb, J.R. Cheeseman et al., Gaussian 09, Polyaniline NH3, CO2, and CO Gas Sensors: Revision D.01,Gaussian, Inc., Wallingford CT, 2009. Comparison with Recent Experimental Data”, J. Phys. Chem. C, vol. 117, pp. 23701−23711, 2013. [17] S.F. Boys, F. Bernardi, “The calculation of small molecular interactions by the differences of separate [23] S. Armaković, S.J. Armaković, J.P. Šetrajčić, L.D. total energies. Some procedures with reduced errors”, Džambas, “Specificities of boron disubstituted Mol. Phys., vol. 19, pp. 553-566, 1970. sumanenes”, J. Mol. Model., vol. 19, pp. 1153–1166, 2013. [18] M.J. Jenita, A. Antony, M. Prabhu, N. Rajendiran, “Theoretical study of inclusion complexation of [24] K.R.S. Chandrakumar, T.K. Ghanty, S.K. Ghosh, tricylic antidepressant drugs with β-cyclodextrin”, “Relationship between ionization potential, Indian Journal of Chemistry, vol. 51A, pp. 1686- polarizability, and softness: a case study of lithium and 1694, 2012. sodium metal clusters”, J. Phys. Chem. A, vol. 108, pp. 6661–6666, 2004. [19] I. Fleming, “Frontier orbitals and organic chemical reactions”, Wiley, London, 1976. [25] R.G. Parr, P.K. Chattaraj, “Principle of maximum hardness”, J. Am. Chem. Soc., vol. 113, pp. 1854– [20] M. Govindarajan, M. Karabacak, A. Suvitha, S. 1855, 1991. Periandy, Spectrochim. Acta A, vol. 89, pp. 137-148, 2012. [26] P.K. Chattaraj, H. Lee, R.G. Parr, “HSAB principle”, J. Am. Chem. Soc., vol. 113, pp. 1855–1856, 1991. [21] M.M. El-Nahass, M.A. Kamel, E.F. El-deeb, A.A. Atta, S.Y. Huthaily, Spectrochim. Acta A, vol. 79, pp. 443- 450, 2011. 206 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 A New Multilevel Inverter Based Parallel Active Power Filter Korhan Karaarslan*‡, Birol Arifoglu*, Ersoy Beser*, Sabri Camur* *Department of Electrical Engineering, Kocaeli University, Umuttepe Campus 41380 Izmit, Kocaeli, Turkey (korhan.karaarslan@kocaeli.edu.tr, barif@kocaeli.edu.tr, ebeser@kocaeli.edu.tr, scamur@kocaeli.edu.tr) ‡ Corresponding Author; Korhan Karaarslan, Department of Electrical Engineering, Kocaeli University, Umuttepe Campus 41380 Izmit, Kocaeli, Turkey, Tel: +90 542 420 2022, Fax: +90 262 303 3003, korhan.karaarslan@kocaeli.edu.tr Received: 14.09.2017 Accepted: 13.12.2017 Abstract- Developments in the field of power electronics have increased the use of non-linear loads. The increase in the use of these loads causes power quality to decrease and problems to appear in transmission and distribution systems. The adverse effect of passive harmonic filters and the up-growing use of modern sensitive loads in the production process have made active power filters an important area of interest in eliminating current and voltage harmonics. In this study, a multilevel inverter topology for parallel active power filter applications has been proposed. The principle of operation for the proposed multilevel inverter topology is explained in detail. The harmonic components of the grid current which has a high total harmonic distortion (THD) value, are eliminated by the proposed multilevel inverter providing a reduced number of switching elements. The generation of switching signals for parallel active power filter application is also expressed with generalized formulas. All required operating states of multilevel inverter topology are given in detail to obtain reference filter current with harmonic components. In addition, the simulation studies carried out for different number of level modules and different values of harmonic distortion are presented. The obtained results show that the proposed multilevel inverter topology is suitable for parallel active power filter applications. Keywords Active power filter, multilevel inverter, total harmonic distortion, power quality. 1. Introduction compensation, resonance suppression and voltage regulation, are mainly classified into two groups according to the grid The widespread use of nonlinear loads such as arc and connection modes; parallel and series active power filters. ladle furnaces, motor drivers, switched-mode power supplies, Parallel active power filters are used to eliminate problems rectifiers and dc/dc converters has increased due to advances created by non-linear loads acting as harmonic current in the field of power electronics. Such loads cause power sources, while series active power filters connected to the quality problems in transmission and distribution systems. system via a transformer provide voltage regulation by Current and voltage harmonics as a problem of power quality eliminating voltage harmonics. become more considerable in parallel with the increasing use The operation principle of the active power filters is based on of modern sensitive loads in the production process [1]. the generation of the reference current and/or voltage Due to the ever-increasing power quality problems, waveforms obtained by a set of control techniques from the active filtering solutions are of interest for harmonic sampled load current and/or voltage with the help of suppression since the conventional passive filters have controlled semiconductor power switches [1]. Based on this adverse effects such that they are affected by the grid operation, it can be said that active power filters consist of impedance and can filter only for the specified harmonic three main parts as shown in Fig.1; harmonic detection unit, component. control unit and inverter unit. Generally, the harmonic detection unit and the control unit are defined together. The Active power filters, which can be applied to solve many harmonic detection unit detects the harmonics from the problems such as filtering harmonics, reactive power 207 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 Level Module 2 V L İL grid f İf L+ H+ Load 2 Q1 Q1 Q1H Q2H İf İL Vs P1 2Vd L- Harmonic N 22 Detection A B Unit Q2L Iref L+1 Q2H Q0 Q0 Control Q0 Q1L Q2LUnit Q0 P1 Microcontroller Vd L- H- N 1 Inverter Unit 1 Fig. 1. The principle scheme of parallel active power filter Level Module 1 H-bridge Module with main parts Fig. 2. The principle scheme for 7-level half-bridge cascaded inverter sampled load current and/or voltage by means of an algorithm thus determining the reference signal. Algorithms applications is the cascaded H-bridge inverter, The reasons developed for the determination of harmonics are carried out for the preference of the cascaded H-bridge inverter are that in this unit. These algorithms are generally divided into two it is suitable for high voltage and high power applications, groups as frequency and time domain based methods. The the dv/dt voltage stress on the switching elements is reduced frequency domain methods used to determine harmonics in and the THD value of the output voltage is lower [8-11]. active power filters are Discrete Fourier Transform (DFT), Multilevel inverters can produce output voltage with Fast Fourier Transform (FFT) and Sliding-Discrete Fourier sinusoidal form as well as output voltage with desired Transform (S-DFT). harmonics. In this way, multilevel inverters are used for The time domain methods applied for the same purpose active power filter applications [11]. In this study, a half- are instantaneous power theory (p-q theory), artificial neural bridge cascaded multilevel inverter topology is proposed. network based techniques, constant reference (αβ) The switching algorithm developed for the proposed transformation, synchronous reference (dq) transformation multilevel inverter based parallel active power filter is also and hybrid (αβ/dq) transformation [2-3]. introduced. The switching signals to obtain the required output waveform and the other results of the filtering process At the input of the current and/or voltage control unit, are demonstrated by simulation studies. The results of the there is the reference signal determined at the harmonic carried out simulation studies confirm that the developed detection unit while the switching signals of the inverter unit strategy works properly. are found at the output. The contribution of this unit to the operating principle of the active power filter is the generation 2. Half-Bridge Cascaded Multilevel Inverter of switching signals for controlled semiconductor power switches. For this purpose, it is seen that different control Figure 2 shows the principle scheme of the 7-level half- techniques such as Pulse Width Modulation (PWM), bridge cascaded inverter to be used for parallel active power Sinusoidal PWM, hysteresis control and recently used "dead- filter application. The inverter is basically composed of two beat" control are used in the literature. different parts, the level module (LM) and the H-bridge The inverter unit is the part where the filter output voltage module (HM) [12]. The required level of the inverter can be (VAB) is obtained according to the reference voltage. The acquired by changing the number of LM. To expand the switching signals from the control unit are used to generate system and increase the output voltage levels of the inverter, the filter output voltage of the inverter. In the literature, more LMs connected in series are used. many studies have been carried out with different inverter LM comprises a dc source and two semiconductor units to reduce the complexity of the control method by switching elements. The voltage of dc source V in LM 1 is reducing the number of semiconductor power switches [4]. d related to the required maximum value of the output voltage When many features such as switching frequency, dv/dt and the level of the inverter. voltage stress on switching elements, efficiency, electromagnetic interference, harmonic distortion and output The voltage of dc sources in other LMs are scaled in filter requirement are taken into consideration, multilevel power of 2. The maximum level of the inverter n can be inverter topologies are often preferred for active power filter computed as applications [5-7]. In the literature, although there are studies on active power filters with diode-clamped and flying- (m 1)n  2 1 (1) capacitor inverters, it has been determined that the most common multilevel topology in active power filter where m is the number of LMs used in the inverter. 208 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 HM shown in Fig. 2 is a conventional H-bridge inverter. Switching signals are generated for the instantaneous HM is the constant part of the multi-level inverter. In order to values of the reference voltage signal Vref. The equation for increase the number of levels, HM is connected to the series- the switching signals used in the system can be generalized connected LMs as shown in Fig. 2. according to the number of level modules as follows; The value of n is the maximum level number of the inverter depending on m calculated by using equation (1).  j  1, V (t) mod 2  ref However, any required level can also be acquired by the    ( j1)  (5) same number of LMs. For example, a number of level V (t)  V (t) mod 2  Q (t)    ref ref   modules of 3, a maximum of 15 levels can be obtained, as ( j1)   j  1,  mod 2( j1)  well as 9, 11 or 13 levels. By increasing the number of LMs,   2     any required number of level can be easily obtained by the   proposed inverter since it has a simple and modular structure. By using equation (5), the switching signals for the 3. Switching Algorithm switching elements used in LMs are easily obtained. There are two switching elements in each level module, and these The multilevel inverter used in the proposed active switching elements are operated as if they are the inverse of power filter has the capability to generate an output voltage each other. waveform more sinusoidally. Furthermore, the multilevel With the obtained switching signals an output voltage inverter can produce any voltage with harmonic components. similar to the reference voltage signal given in Fig.3 can be Due to this feature, it acts as a harmonic voltage source. For produced as shown in Fig.4. this purpose, it is first necessary to define the reference voltage signal which contains the desired harmonic The waveforms of the inverter output voltage given in components [13]. The reference voltage is defined as in Fig.4 can converge to the reference voltage signal by equation (2); increasing the number of level modules. If the inverter output voltage converges to the reference signal, the harmonic V dc  filtering capability of the parallel active power filter V    V sin(ht  ) (2) ref increases. 2 h1 h h The reference voltage signal is present by the harmonic detection unit. A sample reference voltage signal with 3rd and 7th harmonic components is chosen to describe the proposed switching algorithm. The waveform of the sample reference voltage signal is shown in Fig.3, and the mathematical definition of this waveform is given in equation (3). V  V sin(3t  ) V sin(7t  ) (3) ref 3 3 7 7 Fig.3 depicts the sampling time of the switching signals (∆t). The smaller the sampling time, the more similar the output voltage is to the reference voltage signal. The sampling time depends on the frequency of the fundamental (a) 7-level harmonic of the output voltage and the number of level modules. t  t  t  t (4) sample i1 i (b) 31-level Fig. 4. The waveforms of reference voltage signal and output Fig. 3. The waveform of a sample reference voltage signal voltage of a multilevel inverter 209 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 V_grid L V_grid Lf f Load Load Q1 Q1 Q1 Q1 Q1H Q2H Q1H Q2H 2Vd 2Vd Q0 Q0 Q0 Q0 Q1L Q2L Q1L Q2L Vd Vd a) V_grid > 0, VF = 0 e) V_grid > 0, VF = 2Vd V_grid L V_grid Lf f Load Load Q1 Q1 Q1 Q1 Q1H Q2H Q1H Q2H 2Vd 2Vd Q0 Q0 Q0 Q0 Q1L Q2L Q1L Q2L Vd Vd b) V_grid < 0, VF = 0 f) V_grid < 0, VF = -2Vd V_grid L V_grid Lf f Load Load Q1 Q1 Q1 Q1 Q1H Q2H Q1H Q2H 2Vd 2Vd Q0 Q0 Q0 Q0 Q1L Q2L Q1L Q2L Vd Vd c) V_grid > 0, VF = Vd g) V_grid > 0, VF = 3Vd V_grid L V_grid Lf f Load Load Q1 Q1 Q1 Q1 Q1H Q2H Q1H Q2H 2Vd 2Vd Q0 Q0 Q0 Q0 Q1L Q2L Q1L Q2L Vd Vd d) V_grid < 0, VF = -Vd h) V_grid < 0, VF = -3Vd Fig. 5. The operating states of a 7-level parallel active power filter 210 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 Figure 5 shows the topology of the multilevel inverter with the ability to generate the desired voltage waveform at its output through the proposed switching algorithm, including all operating states of a parallel active power filter application. The results of this application are included in the simulation study. 4. Simulation Study (a) The switching algorithm proposed for parallel active power filter is applied to inverters with different number of level modules. For this purpose, a distorted grid current with a THD value of 23.43% is filtered by using 31-level inverter. The grid current consists of 5th, 7th, and 9th harmonic components with fundamental harmonic. After filtering, the grid current is nearly sinusoidal with the THD value of 0.30%. The grid current and its harmonic current (b) components, the grid and filter currents after the parallel active power filter is applied, and switching signals are shown in Fig.6a, 6b and 6c, respectively. To demonstrate the effect of level number, in other words the number of level modules, a highly distorted grid current consisting 3rd and 7th harmonic components except fundamental with a THD value of 34.99% is filtered by using 7-level and 31-level inverter based parallel active power filters. The simulation (c) results are shown in Fig.7 and Fig.8. The grid current and its harmonic current components are shown in Fig.7a and Fig. 7. The simulation results obtained from 7-level inverter Fig.8a; the grid and filter currents after the parallel active based parallel active power filter, a) grid current with a THD power filter is applied are shown in Fig.7b and Fig.8b; and value of 34.99% and harmonic components before filtering, switching signals are shown in Fig.7c and Fig.8c. b) grid current and filter current, c) switching signals (a) (a) (b) (b) (c) (c) Fig. 6. The simulation results obtained from 31-level inverter Fig. 8. The simulation results obtained from 31-level inverter based parallel active power filter, a) grid current with a THD based parallel active power filter, a) grid current with a THD value of 23.43% and harmonic components before filtering, value of 34.99% and harmonic components before filtering, b) grid current and filter current, c) switching signals b) grid current and filter current, c) switching signals 211 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Korhan Karaarslan et al., Vol.3, No.4, 2017 According to the simulation results, it is seen that the [6] M. Ortuzar, R. Carmi, J. Dixon, and L. Moran, grid current after filtering becomes approximately sinusoidal. “Voltage Source Active Power Filter, Based on Multi- Stage Converter and Ultra Capacitor DC-Link,” The 5. Conclusions 29th Annual Conference of the IEEE Industrial Electronics Society (IECON 2003), pp. 2300-2305, Active power filters are frequently used in solving power 2003. quality problems in parallel with the increased use of non- [7] C. Junling, L. Yaohua, W. Ping, Y. Zhizhu, and D. linear loads. In this study, a switching algorithm developed Zuyi, “A Closed-Loop Selective Harmonic for multilevel inverter based parallel active power filter is Compensation with Capacitor Voltage Balancing introduced. Through the switching algorithm used for the Control of Cascaded Multilevel Inverter for High Power acquisition of switching signals, the total harmonic distortion Active Power Filters,” IEEE Power Electronics of 34.99% generated by the third and seventh harmonic Specialists Conference (PESC 2008), pp. 569-573, current component is reduced to 1.70% using two level 2008. modules and 0.27% using four level modules. The results of the simulation studies demonstrate the validity of the control [8] W. Madhukar, and P. Agarwal, “Comparison of Control algorithm and the importance of the number of level modules Strategies for Multilevel Inverter based Active Power used in the inverter unit. Filter used in High Voltage Systems,” Power Electronics Drives and Energy Systems, pp. 1-6, 2010. Acknowledgements [9] Z. Chen, Y. Luo, and M. Chen, “Control and Performance of a Cascaded Shunt Active Power Filter This study has been funded by Kocaeli University. for Aircraft Electric Power System,” IEEE Trans. On Project Number: 060-2014. Industrial Electronics, vol. 59, no. 9, pp. 3617-3623, 2012. References [10] A.A. Valdez-Fernandez, P.R. Martinez-Rodriguez, G. [1] İ. Kocabaş, O. Uçak, and A. Terciyanlı, “DSP Tabanlı Escobar, C.A. Limones-Pozos, and J.M. Sosa, “A Gerilim Kaynaklı Şönt Aktif Güç Filtresi Uygulaması,” Model-Based Controller for the Cascade H-Bridge ELECO 2006, Bursa, Türkiye, pp. 171-175, 2006. Multilevel Converter Used as a Shunt Active Filter,” IEEE Trans. On Industrial Electronics, vol. 60, no. 11, [2] Y.F. Wang, and Y.W. Li, “Three-Phase Cascaded pp. 5019-5028, 2013. Delayed Signal Cancellation PLL for Fast Selective Harmonic Detection,” IEEE Trans. on Industrial [11] E. Beşer, B. Arifoğlu, S. Çamur, and E. Kandemir Electronics, vol. 60, no. 4, pp. 1452-1463, 2013. Beşer, “A Novel Design and Application of A Single Phase Multilevel Inverter,” International Review of [3] Y.F. Wang, and Y.W. Li, “A Fundamental and Harmonic Electrical Engineering, vol. 4, no. 1, pp. 7-13, 2009. Component Detection Method for Single-Phase Systems,” IEEE Trans. on Power Electronics, vol. 28, no. [12] E. Beşer, “Anahtarlama Elemanı Sayısı ve Harmonik 5, pp. 2204-2213, 2013. Optimizasyonu ile Bir Fazlı Çok Seviyeli Evirici Tasarımı,” Kocaeli Üniversitesi Fen Bilimleri Enstitüsü, [4] V. Khadkikar, “Enhancing Electric Power Quality Using Doktora Tezi, Kocaeli, 2009. UPQC: A Comprehensive Overview,” IEEE Trans. on Power Electronics, vol. 27, no. 5, pp. 2284-2297, 2012. [13] E. Beşer, B. Arifoğlu, S. Çamur, and E. Kandemir Beşer, “Design and Application of a Single Phase [5] F. S. Kang, S. J. Park, S. E. Cho, C. U. Kim, and T. Ise, Multilevel Inverter Suitable for Using as a Voltage “Multilevel PWM Inverters Suitable for the Use of Stand- Harmonic Source,” Journal of Power Electronics, vol. Alone Photovoltaic Power Systems,” IEEE Trans. on 10, no. 2, pp. 138-145, 2010. Energy Conversion, vol. 20, no. 4, pp. 906-915, 2005. 212 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 Effects of Arc Voltage and Welding Current on the Arc Length of Tungsten Inert Gas Welding (TIG) Ikpe Aniekan E.*‡, Owunna Ikechukwu**, Ememobong Ikpe*** *‡Department of Mechanical Engineering, University of Benin, Nigeria. **Department of Mechanical Engineering, University of Benin, Nigeria. *** Electrical Electronics/Instrumentation Department, Exxon Mobile Nigeria, Nigeria (ikpeaniekan@gmail.com, ikechukwu.owunna@uniben.edu, ememobong.e.ikpe@gmail.com) ‡ Corresponding Author; Ikpe Aniekan, Room 142, Mechanical Engineering Department, University of Benin, Benin City, Tel: +2349024773812, ikpeaniekan@gmail.com Received: 10.06.2017 Accepted: 13.12.2017 Abstract- The effects of welding current and arc voltage on the arc length of Tungsten Inert Gas (TIG) welding was investigated in this study. Mild steel plates of 10mm thickness and 50mm x 100mm diameter were joined together through TIG welding process for four (4) different intervals and the welding arc lengths were measured consecutively. The measured arc length for each welding interval increased from 1.19mm, 1.93mm, 2.54mm and 3.12mm as the current increased from 50A, 100A, 150A to 190A while the voltage also maintained the same increasing trend from 200V, 240V, 280V to 320V. The minimum arc length values measured for each welding interval with the aforementioned current and voltage range also maintained an increasing trend, and that led to the conclusion that the higher the welding current and arc voltage, the longer and wider the arc length which is an influential factor to other welding attributes. Keywords Welding, Arc voltage, Arc length, Welding current, Electrode, Mild Steel. 1. Introduction Gas (TIG) which is the centre of focus in this study [3]. The above mentioned welding processes has not always been Welding is the process of joining two or more metals widely used in the past but its application and relevance in together [1]. Welding processes can be classified into two many industries in recent times has increased tremendously major groups such as fusion welding and solid-phase welding. due to cost and reliability. Among the aforementioned welding Fusion welding is based on the principles of heat application technologies, TIG welding is gradually emerging as the to join two or more materials together, and it is categorized process of choice for joining thin material, dissimilar and into various types of welding operation which includes arc particularly suitable for welding metals and metal alloys than welding, resistance welding, oxy-fuel welding also referred to any other arc welding process with little smoke or fumes, neat as oxyacetylene welding or oxy welding, electron beam and slag free welds and appearance that may result in ease of welding and laser beam welding. Solid-phase welding which finishing despite its slower travel speeds than other processes, is the other group of welding classification is based on the lower filler metal deposition rates and the cost of equipment application of pressure alone or combination of both heat and which can be higher than using other welding processes [4]. pressure to join two or more materials together, and it is Tungsten inert gas (TIG) welding also referred to as Gas categorized into diffusion welding, friction welding, forge tungsten arc welding (GTAW) is an electric arc welding welding and ultrasonic welding [2]. Arc is divided into four method that applies a non-consumable tungsten electrode in (4) major processes such as Shielded Metal Arc Welding the process of joining two metals together. (SMAW) also known as Manual Metal Arc Welding, Gas Metal Arc Welding (GMAW) also known as Metal Inert Gas Weld penetration is the rate at which the fusion line or Active Gas Welding (MIG/MAG), Flux Core Arc Welding extends below the surface of the welded material. Weld (FCAW) and Gas Tungsten Arc Welding or Tungsten Inert penetration is interrelated with arc length and welding current, 213 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 as increase or decrease in current can result in further increase decreasing the number of droplets [10]. Also, the possibility or decrease in arc length and in the weld penetration depth. of breaking the arc may increase with increasing voltage and Studies have shown that weld penetration is influenced by affecting the welding process. Increasing welding voltage can welding current, polarity, arc travel speed, electrode diameter increase flux consumption which in turn increases or lessen etc. [5]. Memduh et al. [6] noted that weld penetration is the alloying elements, thereby, affecting the mechanical and directly proportional to welding current, and that deep metallurgical properties of the weld metal [11, 13]. In penetration can be achieved in operations where DCEP addition, extremely high voltage has the tendency of polarity is employed whereas, shallow penetration can be producing wider bead geometry that is prone to cracks, achieved as a result of DCEN polarity. Penetration decreases increased undercut and difficulty in slag removal. Low voltage with the increase in welding speed because the time at which produces a stiffer arc, which improves penetration in a deep the arc force is allowed to penetrate into the material’s surface weld groove and resists arc blow, whereas, extremely low decreases. The penetration decreases with the increase in voltage produces a narrow bead geometry and results in electrode diameter due to decrease in current intensity and arc difficulty during slag removal along the bead edges. length [7]. British Standard Institution (BSI) describes arc In a given arc welding operation carried out with constant length as the accumulative distance from the tip of the welding voltage and amperage, the thermal efficiency of heat source is electrode to the adjacent surface of the weld pool [8]. given [14] as shown in equation 2; According to Egerland [9], an average arc length may be expressed by equation 1, 𝑄 𝑡 𝑄𝜂 = 𝑠 𝑤𝑒𝑙𝑑 = 𝑠 (2) 𝑉𝐼𝑡𝑤𝑒𝑙𝑑 𝑉𝐼 𝑙𝑎𝑟𝑐_𝑎𝑣 = 𝑐1 ∗ 𝐿𝑎𝑟𝑐_𝑚𝑎𝑥 + 𝑐2 ∗ 𝐿𝑎𝑟𝑐_𝑚𝑖𝑛 (1) Where, 𝑄𝑠 is the rate of heat generated, tweld is the welding Where, 𝑙𝑎𝑟𝑐_𝑎𝑣 is the average arc length, 𝐿𝑎𝑟𝑐_𝑚𝑎𝑥 is the time and η is the thermal efficiency. maximum arc length, 𝐿𝑎𝑟𝑐_𝑚𝑖𝑛 is the minimum arc length, c1 and c2 are constants showing current correspondence and pulse For welding operations where electric arc serves as the current time. welding source, heat conduction through the workpiece is the primary mode of heat transfer, and the partial differential Tewari et al. [10] studied the effect of welding current, arc equation for transient heat conduction is expressed by voltage, welding speed and heat input rate on the weldability equation 3; of Mild Steel specimens of 50mm× 40mm× 6 mm dimensions 𝜕𝑇 using metal arc welding. The result obtained showed that the 𝜌𝑐 = ∇. (𝑘∇𝑇) + 𝑓 (3) 𝜕𝑡 depth of penetration increased with increasing welding speed up to 110.39 mm/min which was optimum value to obtain Where p is the density, c the specific heat and k is the thermal maximum penetration. Therefore, increasing the speed of conductivity which are all dependent on the temperature T. t travel and maintaining constant arc voltage and current is the welding time and f is the additional heat generated in the resulted in increased penetration until optimum speed was workpiece. Considering the heat generated as a result of arc reached at which penetration was maximum. Increasing the welding process, the heat flux vector is shown in equation 4; speed beyond this optimum resulted in decreased penetration. ?⃗? = −𝑘∇𝑇 (4) Ghazvinloo et al. [11] reported that if the welding speed decreases beyond a certain point, the penetration also will The relationship between enthalpy and the temperature is decrease due to the pressure of the large amount of weld pool expressed by equation 5; beneath the electrode, which will cushion the arc penetrating 𝑇 force. However, high welding voltage produces wider, flatter 𝐻 = ∫ 𝑐(𝜏) 𝑑𝜏 (5) 𝑇𝑟𝑒𝑓 and less deeply penetrating welds than low welding voltages. Depth of penetration is maximum at optimum arc voltage [7]. This implies that; Welding voltage varies with the arc length between the 𝑑𝐻 𝑐 = (6) welding electrode and molten weld metal. An increase in arc 𝑑𝑇 length oftentimes lead to increase in the arc voltage because From equation (3) and (6), the apparent heat capacity equation extension of the arc exposes the entire arc column to the cool can be expressed by equation 7; boundary of the arc [12]. Similarly, the arc column continues to lose the charge carriers through radial movement into the 𝜕𝐻 𝜌 = ∇. (𝑘∇𝑇) + 𝑓 (7) cool boundary of the arc, thereby, introducing a higher 𝜕𝑡 requirement for maintaining adequate charge carriers between Welding current is the electrical amperage in the power the welding electrode and the weld material [5]. The shape of equipment used in carrying out welding operation. It is usually the weld bead cross section and its physical appearance is read from the power meter, but depending on the connection mainly determined by the voltage. Furthermore, if the welding could have a separate ammeter. Welding current is one of the voltage is increased with constant welding current and most influential welding parameter because it has high welding speed, a flatter, wider, and less penetrated weld beads tendency of affecting bead geometry, control the rate at which may be obtained which tends to reduce the porosity due to electrode is melted, controls deposition rate, heat affected corrosion or scale on steel materials. Increasing arc voltage zone, penetration depth and the amount of base metal melted can equally lead to increase in the size of droplets as well as depending on how the amperage is regulated during welding increase in the droplet transfer movement time and thus, operation [15]. 214 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 If the amps is too low, may result in a tall, narrow bead lacking in penetration. The weld will be difficult to start and the arc prone to straying towards one side of a joint in preference to the other. In cases where the bead is too high, the bead has a wider appearance that is flat and irregular, and a small undercut can observed on the right hand side of the weld in the other sectioned of the image below. A deep crater forms at the end of the weld, and removing the slag from the edges of the weld may be difficult. Hence, excessive current should not be compensated by excessive travel speed [16]. This has a high tendency of resulting in slag inclusions due to rapid cooling of the weld. Moreover, with the amps correctly adjusted, a bead of smooth round shape is obtained, and the slag can be removed easily. The increase in electrode melting rate in kg/min as a result of increase in welding current is given by the relationship in equation 8, Fig 2. Workpiece welded with 100A Current Electrode Melting Rate = 1 𝑑2 𝐼𝐿 x 25.4 1.22 [0.35 + + 2.08 x 10−7 x ( 2 ) ] (8) 1000 645 𝑑 Where, d = thickness of the melted electrode L = length of the melted electrode I = welding current. 2. Methodology 5 pieces of mild steel plate with 10mm thickness and dimension of 50mm x 100mm each were grinded to smoothen the surfaces and joined together in series of five (5) different welded grooves (joints), and the arc length which contributes to the depth of weld penetration was measured during the welding operation. Using TIG welding machine with AC currents of 50A and voltage ranges of 200V, 240V, 280V and 320V; 100A with the same voltage ranges; 150A with the Fig 3. Workpiece welded with 150A Current same voltage ranges and 190A with the same voltage ranges on each welded plate; the procedure was carried out on four (4) different 10mm mild steel plate of the same specification. The welding set-up consisted of speed control unit, TIG welding touch and clamp, tungsten electrode, gas cylinder containing argon gas. Several welding variations noted during the welding experiment are reported in this study. Figure 1-4 represent the four (4) mild steel plates used for the experiment. Fig 4. Workpiece welded with Current of 190A 3. Results Results obtained from the welding conditions and process parameters stated earlier in the methodology are presented as follows; Fig 1. Workpiece welded with 50A Current 215 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 Table 1. Result of Welding Arc Length obtained with 50A and variable Voltage Ranges Arc Voltages (V) Welding Current (A) 200V 240V 280V 320V Arc Lengths (mm) 0.62 0.92 1.4 1.19 0.68 0.83 1.1 1.11 50 A 0.52 0.96 1.3 1.5 0.49 0.72 1 1.12 0.68 0.81 1.4 1.16 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0 1 2 3 4 5 200V 240V 280V 320V Fig 5. Trend of Arc length with 50A and Variable Voltage ranges Fig 6. Measurement of Welding Arc Length with 50A and Variable Voltage Ranges Table 2. Result of Welding Arc Length obtained with 100A and variable Voltage Ranges Arc Voltages (V) Welding Current (A) 200V 240V 280V 320V Arc Lengths (mm) 1.24 1.32 1.58 1.88 1.28 1.38 1.43 1.62 100 A 1.2 1.36 1.49 1.93 1.27 1.29 1.57 1.74 1.25 1.34 1.46 1.86 216 Arc Length (mm) INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 2,5 2 1,5 1 0,5 0 1 2 3 4 5 200V 240V 280V 320V Fig 7. Trend of Arc length with 100A and Variable Voltage ranges Fig 8. Measurement of Welding Arc Length with 100A and Variable Voltage Ranges Table 3. Result of Welding Arc Length obtained with 150A and variable Voltage Ranges Arc Voltages (V) Welding Current (A) 200V 240V 280V 320V Arc Lengths (mm) 1.51 1.72 2 2.51 1.54 1.75 2.1 2.22 150 A 1.59 1.54 1.92 2.32 1.4 1.63 2.19 2.34 1.46 1.6 1.84 2.54 217 Arc Length (mm) INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 3 2,5 2 1,5 1 0,5 0 1 2 3 4 5 200V 240V 280V 320V Fig 9. Trend of Arc length with 150A and Variable Voltage ranges Fig 10. Measurement of Welding Arc Length with 150A and Variable Voltage Ranges Table 4. Result of Welding Arc Length obtained with 190A and variable Voltage Ranges Arc Voltages (V) Welding Current (A) 200V 240V 280V 320V Arc Lengths (mm) 2.33 2.63 2.72 2.78 2.37 2.42 2.74 3.11 190 A 2.26 2.54 2.63 2.91 2.34 2.66 2.73 2.82 2.28 2.49 2.68 3.12 3,5 3 2,5 2 1,5 1 0,5 0 1 2 3 4 5 200V 240V 280V 320V Fig 11. Trend of Arc length with 190A and Variable Voltage ranges 218 Arc Length (mm) Arc Length (mm) INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 Fig 12. Measurement of Welding Arc Length with 190A and Variable Voltage Ranges Fig 13. Maximum Arc length Measured for current applied in each of the Four (4) Welding Categories 4. Discussion Figure 1-4 indicates behaviour of the welded material white spatters were found especially on the weldment with under different current and voltage ranges. It can be observed few more on the base metal. The spattering effect intensified in Fig 1 that using a current of 50A with voltage ranging from in Fig 4 where current of 190A with voltage ranging from 200- 200-320, only a handful of spatters were found on the base 320. The increasing number of spatters which rather produced metal and the weldment which is quite acceptable. By roughness surfaces was as a result of increase in current input definition, spattering in welding application implies having and vice versa. Table 1 represents the measured welding arc tiny droplet of electrodes on or beside the weldment and may length obtained from a fixed current of 50A and variable either fuse on them. In Fig 2 where current of 100A with voltage ranging from 200-320. It is obvious that with a fixed voltage ranging from 200-320 were used, the base metal and current and increasing voltage, the arc length is likely to the weldment is smoother with little or no spatters on the increase as well. This can be observed with full clarity in Fig surface. However in Fig 3 where current of 150A with voltage 5 showing the trend of Arc length with a fixed current (50A) ranging from 200-320 were used, microscopically black and and variable voltage ranges (200V, 240V, 280V and 320V), 219 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 and in Fig 6 depicting the measured welding arc length with ii. The maximum arc length measured was 3.12mm with the same fixed current (50A) and variable voltage ranges current amperage of 190A and arc voltage input of 320V. (200V, 240V, 280V and 320V) respectively. The same trend of increase was likewise observed for subsequent fixed iii. Slight change in arc length caused significant change in currents of 100A, 150A and 190A and similar voltage ranges current therefore, electrode melting rate and metal (200V, 240V, 280V and 320V) as shown in Table 2-4 and Fig deposition changed rapidly in response. 7-12 respectively. Fig 13 depicts the maximum Arc length iv. A change in welding current and arc voltage caused measured for 50A, 100A, 150A and 190A applicable to each corresponding change in arc length. of the four (4) welding categories. The maximum measured arc lengths were derived from the welding operation which v. Lengthening the arc exposed the arc column to cool also correlates with some of the arc length discussed by boundary with more heat losses, thereby, necessitating Egerland [9]. high voltage requirement. Although the welding arc penetration was not considered vi. Longer arc length produced unstable welding arc, reduced in this study, several existing studies have shown that arc penetration, increased spatter, flatter and wider welds, and penetration oftentimes increase with the increase in welding prevents shielding gas from protecting the molten puddle current. In other words, too high a welding current at a given atmospheric contamination. welding speed can influence the depth of fusion or penetration by making it so high that the resultant weldment tends to melt vii. Long arc length increased puddle heat, produced flatter through the metal being welded. High current application in welds, increased penetration, reduced spatter, produced welding operation may result in a waste of electrodes in the stable welding arc. form of excessive reinforcement and can as well produce viii. Short arc length produced flatter welds, less puddle heat, digging arc and undercuts which increases distortion in less penetration. weldment and shrinkage. As the welding current increases, the bead width also increase until a critical point is attained and ix. Shorter arc length produced less puddle heat, high then gradually decreases if polarity used is that of Direct tendency of electrode to stick, poor penetration, uneven Current Electrode Positive (DCEP). Also, if the polarity used beads with irregular ripples. is that of Direct Current Electrode Negative (DCEN), the bead width increases as current for entire range increases. For the References same flux, HAZ also increases with increasing welding current. [1] S. S. Pawale, Theoretical & experimental study of MIG/MAG welding technique. International Journal of However, if current used for the welding operation is too Engineering Trends and Technology (IJETT), Vol. 24, low, inadequate penetration or incomplete fusion may occur, Issue 3, pp. 142-144, 2015. and too low a welding current may equally result in inadequate penetration, unstable arc and overlapping [17]. Welding [2] A. O. Akii Ibhadode, Introduction to Manufacturing current also influences the heat requirement for desired weld Technology, Second Edition, Benin City: Ambik pool and fusion of similar and dissimilar metals. The area of Publishers, ISBN: 978-8016-34-0, 2001, pp. 553-555, heat incident on the workpiece surface determines the heat 2001. distribution on metal surface [18]. The arc behaviour varies depending on whether the electrode is connected to the [3] K. Weman, Welding processes handbook, ISBN positive or negative terminal of the power source. With 0849317738, New York: CRC Press LLC, 2003. electrode positive (d.c.+) polarity, the electrode generates greater output heat but less penetration during welding [4] Miller Electric, Guidelines for Gas Tungsten Arc Welding operation than with electrode negative (d.c.−) polarity. The (GTAW), Illinois Tool Works Company, Appleton, USA, current-carrying capacity of an electrode of a given size will Vol. 215 994 D, pp. 5, 2013. therefore be lower with positive polarity than with negative [5] S. Mohd, P. Mohd and K. Pratibha, Effect of MIG Welding polarity [19]. Jatinder and Jagdev [20] studied the effect of Input Process Parameters on Weld Bead Geometry on welding speed and heat input rate on stress concentration HSLA Steel. International Journal of Engineering Science factor of butt welded joint of IS 2062 E 250 A steel plates and Technology, Vol. 5, Issue1, pp. 200-212, 2013. using GMAW and observed that with increase in the heat input rate, flank angle and weld bead width, reinforcement height [6] K. Memduh, Y. Ahmet, B. Mustafa and C. Zarif, Effects increases, whereas notch radius decreased with an increase in of Welding Current and Arc Voltage on FCAW Weld Bead the stress concentration factor increases. Geometry. International Journal of Research in Engineering and Technology Vol. 4, Issue 9, pp. 23-28 5. Conclusion 2015. The following conclusions were dawned from the [7] P. C. Satyaduttsinh, V. Jayesh and M. Tushar, A Review experimental approach conducted in this study; on Optimization of MIG Welding Parameters using Taguchi’s DOE Method. International Journal of i. The measured minimum arc length was 0.49mm with Engineering and Management Research Vol. 4, Issue 1, current of 50A and arc voltage of 200V. pp. 16-21, 2014. 220 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Ikpe Aniekan E. et al., Vol.3, No.4, 2017 [8] BIS-British Standards Institution, British standard BS 499- Plates. Indian Journal of Engineering and Material 1:2009: welding terms and symbols part 1: glossary for Science Vol. 18, pp. 425-434, 2011. welding, brazing and thermal cutting, BSI London, 2009. [15] L. L. Wang, F. G. Lu, H. C. Cui and X. H. Tang, [9] S. Egerland, A Contribution to Arc Length Discussion, Investigation on Thermal Inertia of GMAW-P Welding Soldagem & Inspecao, Fronius International GmbH, Wels, on Al Alloy. Journal of Science and Technology of Austria, Vol. 20, Issue 3, 2015. Welding and Joining Vol. 20, Issue 2, pp. 106-114, 2015. [10] S. P. Tewari, A. Gupta, and J. Prakash, Effect of Welding [16] S. Krishnan, D. V. Kulkarni and A. De, Multipass Pulsed Parameters on the Weldability of Material, International Current Gas Metal Arc Welding of P91 Steel. Journal of Journal of Engineering Science and Technology Vol. 2, Science and Technology of Welding and Joining Vol. 21, Issue 4, pp. 512-516, 2010. Issue3, pp. 171-177, 2016. [11] H. R. Ghazvinloo, A. Honarbakhsh-Raouf and N. [17] M. Weglowski, Y. Huang and Y. M. Zhang, Effect of Shadfar, Effect of arc voltage, welding current and welding current on metal transfer in GMAW Archives of welding speed on fatigue life, impact energy and bead Materials Science and Engineering Vol. 33, Issue 1, pp. penetration of AA 6061 joints produced by robotic MIG 49-56, 2008. welding, Indian Journal of Science and Technology Vol. 3, Issue 2, pp. 156-162, 2010. [18] R. Chandra, S. Arul, and R. Sellamuthu, Effect of Electrode Diameter and Input Current on Gas Tungsten [12] F. Q. Fan, Z. D. Zhang and L. M. Liu, Oxide Contribution Arc Welding Heat Distribution Parameters. International on Arc Plasma in Tungsten Inert Gas Welding of Conference on Advances in Manufacturing and Magnesium Alloy. Journal of Science and Technology Materials Engineering, Procedia Materials Science Vol. of Welding and Joining Vol. 18, Issue 5, pp. 434-440, 5, pp. 2369-2375, 2014. 2013. [19] American Welding Society, Specification for Tungsten [13] Y. T. Lin, D. P. Wang, M. C. Wang, Y. Zhang and Y. Z. and Oxide Dispersed Tungsten Electrodes for Arc He, Effect of Different and Pre-and Post-Weld Heat Welding and Cutting, AWS A5.12M/A5.12:2009 (ISO Treatment on Microstructures and Mechanical 6848:2004 MOD), 550 N.W. LeJeune Road, Miami, FL Properties of Variable Polarity TIG Welded AA2219 33126, 2009. Joints. Journal of Science and Technology of Welding and Joining Vol. 21, Issue 3, pp. 234-241, 2016. [20] G. Jatinder, and S. Jagdev, (2012) Effect of Welding Speed and Heat Input Rate on stress concentration Factor [14] M. Jeyakumar, T. Christopher, R. Narayanan and B. N. of Butt Welded Joint of Is 2062 E250 A Steel. Rao, Residual Stress Evaluation in Butt-Welded Steel International Journal of Advanced Engineering Research and Studies Vol. 1, Issue 3, pp. 98-100, 2012. 221 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 Interactive Fuzzy Decision Making Algorithm for Two Level Linear Fractional Programming Problems Hasan Dalman* *Department of Computer Engineering, Faculty of Engineering and Architecture, Istanbul Gelişim University, Avcilar, 34510, Istanbul, Turkey (hdalman@gelisim.edu.tr) ‡ Corresponding Author; Hasan Dalman, Department of Computer Engineering, Faculty of Engineering and Architecture, Istanbul Gelişim University, Avcilar, 34510, Istanbul, Turkey, Tel: +90 507 723 3421, hdalman@gelisim.edu.tr Received: 26.10.2017 Accepted: 13.12.2017 Abstract- This paper offers an interactive fuzzy decision-making algorithm for solving two-level linear fractional programming (TLLFP) problem which contains a single decision maker at the upper level and multiple decision makers at the lower level. In the presented interactive mechanism, the fuzzy goals and associated weight of the objective at all levels are first determined and the satisfactory solution is attained by renewing the satisfactory degrees of decision makers including the overall satisfactory balance among all levels. Moreover, the value of distance function is used in order to verify the satisfaction grades. Finally, a numerical example is given to illustrate the performance of the presented algorithm. Keywords Two level linear fractional programming problem; fuzzy programming; fuzzy goals; Interactive methods. 1. Introduction descent method (Savard and Gauvin, [9]). and genetic algorithm (Liu, [10]). A fuzzy multilevel programming Multilevel programming problems usually occur in a model is presented by Gao and Liu [11]. They defined a much hierarchical system of large organizations such as Stackelberg–Nash equilibrium. These classical methods are government offices, profit or non-profit organizations, based on Karush–Kuhn–Tucker conditions and/or penalty manufacturing plants, logistic companies, etc. Solution functions [12]. Furthermore, the Stackelberg method does procedures show that all Decision Makers (DMs) has a single not provide Pareto optimality because of its non-cooperative objective, a set of decision variables and a set of general nature [13]. These solution procedures are related to Karush– constraints that affect all DMs. Each unit individually Kuhn–Tucker conditions and/or penalty functions [12]. searches itself earnings. But each of them is affected via the Besides, solution procedure of the Stackelberg method does actions of other units. not give Pareto optimality because of its noncooperative structure [13]. In such hierarchical decisions, it has been Multilevel programming proposed by Bracken and concluded that each DM should have a difficulty of McGill [1] to model a decentralized noncooperative decision motivation to cooperate with the other, and a minimum level system with one leader and multiple followers in 1973. of satisfaction of the DM at a lower level must be subject to Multilevel programming is an NP-hard problem [5]. The the overall profit of the organization. In order to satisfactory Stackelberg method has been employed to solve the solutions, fuzzy set theory to multilevel programming multilevel programming problems. It has much applicability problems was first applied by Lai [12] in 1996. By utilizing in practical such as strategic planning (Bracken and McGill, a search procedure and fuzzy set theory, this procedure of [2]), resource allocation (Aiyoshi and Shimizu, [3]), and satisfactory solution was improved by Shih et al. [14, 15]. water management (Anandalingam and Apprey, [4]). In Moreover, fuzzy programming approaches were employed order to establish mathematical model of multilevel by many authors for solving multiple level linear programming, many methods and algorithms have been programming problems [16, 17], bilevel quadratic fractional proposed such as extreme point algorithm (Candler and programming problem [13, 17, 18], two-level non-convex Towersley, [6]), k.th best algorithm (Bialas and Karwan, programming problems with fuzzy parameters [18], [7]), branch and bound algorithm (Bard and Falk, [8]), 222 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 decentralized two-level linear programming problems [19, its objective function and is affected by the activities of the 20] and so on. other DMs. A mathematical model of two level linear fractional Recently, Baky [21] presented two fuzzy goal programming problem is formulated as follows (see; programming algorithm for multi-level multi-objective linear Ahlacioglu and Tiryaki [13]): programming problems. Arora and Gupta [22] presented an c x  interactive fuzzy goal programming algorithm connecting Upper Level: max f0 x  0 0 x0 bilevel programming problems with the theory of dynamic d0x  0 programming. Wang et al. [23] introduced a concept to deal c x Lower Level: max f x  i i ,i 1,2,...,k (1) i with the bilevel multilevel programming problem. A fuzzy xi dix  i TOPSIS algorithm is introduced in [24]. The distance Subject to function, which was introduced by Yu [24], has been widely xS x n A x  b,x  0 employed to obtain compromise solutions of multi objective programming problems. Moitra and Pal [25] applied fuzzy goal programming method with the theory of distance where; c0  c00 ,c01,c02 ,...,c0k ,ci  ci0 ,ci1,ci2 ,...,cik , function and produced a satisfactory balance by lessening the d0  d00 ,d01,d02 ,...,d0k  and deviations of the leader and follower as far as for bilevel programming. However, some interesting interactive fuzzy di  di0 ,di1,di2 ,...,dik ,i 1,2,...,k are n  dimensional fixed decision making algorithms have widely been employed to row vectors;  , , and  ,i 1,2,...,k are reel numbers; 0 0 i i obtain the efficient results of bilevel and multilevel b is m  dimensional constant column vector; A is an mn programming problems [27, 28, 29, 30, 32]. Toksari and constant matrix with full rank r. S is a non-empty, convex Bilim [31] introduced an interactive fuzzy goal method based and compact set in n ; d x   and d x  are greater on the Jacobian matrix for solving the multilevel fractional 0 0 i i programming problem. than zero. An interactive fuzzy decision making algorithm in this 2.1. Construction of Membership Function paper is presented for two-level linear fractional Each of the decision makers aims to minimize its own programming problems (TLLFPP) a single DM at the first objective over the feasible region. The optimal solutions of level as well as multiple decision makers at the second level. them are found, individually and these solutions can be Objective functions and constraint functions for DMs at both chosen as the best solution. Besides, the achieved value of levels are fractional and linear functions, respectively. In each of objective can be admitted as the aspiration levels for order to solve the problem, the fuzzy goal of each of the corresponding fuzzy goals. For convenience, the method objective function is defined by getting individual optimal given in paper [28] is used to determine membership solutions. Thereafter, the membership function of each functions. Let us   f , j  0,1,2,...k to define the fuzzy goals fractional objective for TLLFPP is constructed. Then the j overall satisfactory balance between the leader and the of the leader and the follower, respectively. follower is defined by introducing a new balance function. 0 f  f L Finally, numerical examples are presented to demonstrate the  if j j  f L j  f j L U feasibility of the presented interactive algorithm.   f j    if f j  f j  f j (2)  f U j  f L j if U The remaining of this paper is arranged as follows. A  f j  f j 1 mathematical model of bilevel fractional programming where f U is called an ideal value and f L is tolerance limit problem is given in Section 2 and an interactive fuzzy j j U L decision making algorithm is presented in Section 3. At least, of j  the fuzzy goal. f j and f j denote the values of the two comparative examples are implemented in Section 4 and objective f j x, j  0,1,2,...k such that the degrees of the the paper is concluded in Section 5. membership function are 1 and 0, respectively. For the sake 2. Problem Formulation of simplicity, we suppose that f U j and f L j are the optimal solutions of the following fractional problems, respectively. In (TLLFPP), two DMs are located at two diverse For instance, hierarchical levels including multiple objectives. Moreover, f Uj  max f j x, j  0,1,2,...k (3) each DM independently controls a set of decision variables. and The first level decision maker (DM) is known as the leader, f Lj  min f j x, j  0,1,2,...k (4) which executes its decision in the scope of the second level DMs known as the follower. Here, each DM tries to optimize 223 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 Start Solve problem (1) as a single objective, individually. Determine equations (3) and (4). Then, compute their own weights for each level using equation (8), respectively. Define the membership functions (2). Construct balance function (6) to estimate the overall satisfactory degree insert the initial minimal acceptable satisfactory levels of leader and follower The leader or/and the follower reduces Formulation of fuzzy programming model (7) his/her or/and their minimal acceptable satisfactory levels Solve fuzzy programming model (7) Yes Does not there exist a solution to (7) ? The leader and the follower update the minimal acceptable satisfactory levels No Calculate the value of distance function and ratio of satisfactory degree (5) is the leader satisfied by the solution? No Yes The solution is the satisfactory efficient solution for leader and follower Fig. 1. Application framework of the interactive fuzzy decision making algorithm 3. Interactive Fuzzy Decision Making Method membership function for a fuzzy goal is forever desired by a DM. But it is difficult to obtain the highest degree for all In decision making process, achievement value of highest membership function values. Therefore, we need the theory 224 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 of the general satisfactory degree between the leader and Proof: If x*, y*  is not an efficient solution, then there exists follower. To do this, the following concept is given [12]: k x , y S such that f * *j x , y   f j x , y  for all wi  fi    i1 (5) j  0,1,2,...,k and fk x , y   f x * k , y *  , j  k for at least one w0  f0  index k. This contradicts that x*, y*  is an optimal solution of Note that, many authors implemented the notation given in above to achieve the satisfactory degree between the leader (6). and follower at the decision making process [12, 22]. When the leader achieves the solution of problem (7) as a Nevertheless, this condition may result in some efficient satisfactory solution, the iterative process finishes. Now, we computation. But, it can be inadequate for complex and consider the following idea for refreshing the minimal large-scale calculations. Therefore, the following balance acceptable satisfactory level  (see page 92 of [34]): 0 function is presented to estimate the overall satisfactory If the leader is not satisfied with the achieved solution and degree which can be characterized as the ratio of two experts that it is desirable to increase the satisfactory degree functions: of the leader at the expense of the satisfactory degree of the k follower, then he/she increases the minimal acceptable 2  f j x  f Lj  satisfactory level  . Otherwise, if the leader experts that it 0 j0 d x  (6) k k is desirable to increase the satisfactory degree of the follower 2 2  f j x  f Uj   f U Lj  f j  at the expense of the satisfactory degree of the leader, then j0 j0 he/she decreases the minimal acceptable satisfactory level Clearly, 0  d  x 1 for all xS.  1. If each decision maker achieves the ideal value, d  x is equal 3.1. The proposed Interactive Fuzzy Decision Algorithm to Solve TLLFPP to 1. In addition, d  x grows as the objective function values  Step 1 Solve the problem (1) as in equation (3) and of the leader and the follower are regenerated. Therefore, we (4) by taking single objective function at a time and can employ the value of d  x to balance the overall neglecting all others. satisfactory degree between the leader and follower at the  Step 2 Determine the ideal values f Uj ( j  0,1,2,...,k ) two level decision making process. L Now, the formulation of the proposed method can be stated and tolerance limits f j ( j  0,1,2,...,k )and weights w for all i as: objective. max d  x  Step 3 Construct the membership functions (2) and subject to then combine all of objective with their own weights,   f0   0 (7) respectively. w0  Step 4 Construct balance function (6). k   fi    1  Step 5 The leader and follower insert the initial i1 wi minimal acceptable satisfactory levels  , . xS 0 1 where  and  are the minimal acceptable satisfactory 0 1  Step 6 Formulate the fuzzy decision making levels specified by the leader and the follower, respectively. programming model (7). Then, solve problem (7) to obtain w and w ,i 1,2,...,k are importance weight of each the optimal solutions. 0 i objective. Here, lover level functions are combined using  Step 7 If there does not exist a solution to (7), the their own weights. leader or/and the follower reduces his/her or/and their Furthermore, w is always 1 for single decision at the upper 0 minimal acceptable satisfactory levels, until a solution level. Therefore, the weights are not considered and by x*, y*  is obtained for (7). taking into account the minimal satisfactory level of the objective at the upper level and by determining a  Step 8 If the leader is satisfied by the solution in proportional satisfaction balance among all objectives and Step 7, go to Step 9, else go to Step 10. their importance weights, we aim to achieve a satisfying  Step 9 The solution is the satisfactory efficient solution from a Pareto optimal solution set for TLLFP solution for leader and follower in problem (1). problem such that the satisfactory levels of all objectives are proportional to their own weights. Here, w ,i 1,2,...,k is  Step 10 The leader and the follower update the i calculated as follows: (see: Kassem [33]): minimal acceptable satisfactory levels  and  , go to step 0 1 f L  f U (6). w  i ii , i 1,2,...,l. (8) l  f L  f U A comparison of results based on linearization procedures i i k 1 given above is shown in Fig. 1. Theorem: If x*, y*  is an optimal solution to problem (7), In order to evaluate the satisfaction, we not only use the then it is also an efficient solution to problem (1). value of the overall satisfactory degree d  x , but also the 225 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 2  k 2 0.1343x01  5x02  x11  3x12  x  21  2x22  x31  2x32  value of distance function D    0.671  1  f  j   (for details,  x01  2x02  2x11  x12  x22  x31  2x32 1 j0  Lover level membership functions: see; [35, 36] )where   f j  represents the achieved 0 if f  0 membership value of the 1j  th decision maker.  f  0   f1   1  if 0  f  0.765 1 0.765  0 if 4. Numerical Example f1  0.7651 1.307x11  x12  The suggested interactive fuzzy decision making method   f1  2x  x 1 will be used to a known numerical example. The following 11 12 numerical example was given by Ahlatcioglu and Tiryaki [13]. They used the decentralized method to solve the 0 if f2  0.804 following problem.  f  0.804   f2   2  if 0.804  f2  0.849 0.849  0.804 Example if f2  0.849 1 Upper level: 0.605x21  x 3x01  5x02  x11  3x12  x21  2x22  x 22 max f  31  2x32    f    0.486 2 0 x  x 1 x0 x01  2x 21 22 02  2x11  x12  x22  x31  2x32 1 Lower level: 0  if f3  2.16 x  f3  2.16 max f  11  x12 x  x x  2x  3 1 ,max f  21 22 ,max f  31 32   f3    if 2.16  f3  5 2 3 x1 2x x2 x  5  2.16211  x12 1 x21  x22 1 x32 1 if f3  5  1 g  2x  x 0.375x  2x  31 01 02  x11  x12  2x21  2x22  x31  3x32 12, 31 32  f   0.873   3  g2  x01  x02  2x11  4x12  3x21  x22  2x31  x32  24, x32 1 g3  3x01  x  3x  x  x  2x  9, and all of them are combined using their weights as: 02 11 21 31 32  g  x 1.307x  x  0.605x  x  0.375x31  2x  3 11 12 21 22 32 4 01  2x02  5x12  x22  2x31  x32 10,  0.486  0.873  3   fg  x i  2x11  x12 1 x21  x 1 x22 32 1  5 01  x02  x11  4x21  5x22  x32  20,     w 0.172 0.371 0.600 g6  4x01  3x02  2x11  x12  x21  x  36, i1 i 22  Step 4: s.t.g7  5x 01  2x02  x12  3x22  4x32  30,  Balance function (6) is determined as g8  2x01  x02  5x11  x21  2x 32  20, 2 g9  x01  x02  2x11  x22  4x  x  8,  3x01  5x02  x11  3x12  x21  2x22  x31  2x 32 31 32   5  g10  x01  3x11  2x12  x21  4x  2x  48,  x01  2x02  2x11  x12  x22  x31  2x32 1  31 32  2 2g11  3x01  5x11  2x21  5x22  2x32 15,  x  11  x12 x21  x   22     0.804 g12  x01  2x02  x  x  3x  5x  x  60  2x11  x12 1  x21  x22 111 12 21 31 32   x01, x02 , x11, x12 , x21, x22 , x31, x32  0  x31  2x32  3   2.16 where x  x , x x 10 01 02 , x1  x11, x12 , x2  x21, x22  and  32  d x  2 x3  x31, x32 .  3x01  5x02  x11  3x12  x21  2x22  x31  2x 32   2.455 Table 1 presents the individual minimum and maximum  x01  2x02  2x11  x12  x22  x31  2x32 1  values (Step 1), the ideal values, tolerance limits and weights 2 2  x  x   x  x  (Step 2) of all the objective functions in both the levels.  11 12  0.765  21 22    0.849 Step 3:  2x11  x12 1   x21  x22 1  Upper level membership function:  x   31  2x32  3   5  66.024 0 x 1 if f  5  32  0  f  5   f   0 if 5  f  2.455 Step 5: Let  1 and 0 1  0.9. 0 0 2.455  5 if f 1 0  2.455  Table 1: The individual minimum and maximum values, the ideal value and tolerance limits and weights f f1 f 0 2 f 3 max f j 2.455 0.765 0.849 5 min f j 5 0 -.804 2.160 f Uj 2.455 0.765 0.849 5 f Lj 5 0 -.804 2.160 226 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 w j 1 0.172 0.371 0.600 Step 6: Then, the corresponding problem (7) can be The ratio of satisfactory degrees is   0.934. formulated as: We execute a comparison with the obtained solutions from 2  [13] in Table 2. From the obtained solutions of d  x and D , 3x01  5x02  x11  3x12  x21  2x22  x31  2x 32   5  x01  2x  2x  x  x  x  2x 1 the obtained solution of the suggested method in this paper is 02 11 12 22 31 32  2 2 better than the method of Tiryaki and Ahlatcioglu[13].  x     11  x12 x21  x 22  0.804 Furthermore, all of the sum of the leader’s values and the      2x11  x12 1  x21  x22 1  follower’s values generated by our suggested method is  x  2x  3  greater than that generated by Ahlatcioglu and Tiryaki [13].  31 32  2.16 So, these solutions indicate that the suggested method in this  x32 1  max d x  paper is practicable. 2  3x  5x  x  3x  x  2x  x  2x  Numerical results prove that the suggested method in this 01 02 11 12 21 22 31 32   2.455  x  2x  2x  x  x  x  2x 1  paper has the following interesting features. 01 02 11 12 22 31 32 2 2  According to Table 2, we can observe that the value  x11  x    12 x  x   0.765 21 22     0.849 of D by the suggested method is smaller than that  2x11  x12 1   x21  x22 1  of other method.  x  2x  3  It should be noted that the larger value of  in (5) is not the  31 32  5  66.024 x 1 more satisfactory the solution. We can see from the distance  32  function D. 5. Conclusion Subject to In this paper, a new interactive fuzzy decision making 0.1343x01  5x02  x11  3x12  x21  2x22  x31  2x32  method based on the idea of the membership function is  0.6711, x  2x  2x  x  x  x  2x 1 suggested for solving the two-level fractional programming 01 02 11 12 22 31 32       problem. We use the overall satisfactory balance between the 1.307 x  x 0.605 x  x 0.375 x11 12 21 22  0.486 31  2x32  3  0.873 leader and the follower into consideration by introducing a 2x x 111  x12 1 x  x 21 22 1  32  0.9, new balance function. Then, a satisfactory solution is 0.172 0.371 0.600 achieved. This solution involves knowledge concerning g1  2x01  x02  x11  x12  2x21  2x22  x31  3x32 12, importance weights of lower level objectives and the g2  x01  x02  2x11  4x12  3x21  x22  2x31  x32  24, minimal satisfactory level of all objectives. Furthermore, this g  3x  x  3x method has an interactive structure as it provides leader to 3 01 02 11  x21  x31  2x32  9, g  x  2x  5x  x  2x  x 10, provide the opportunity of exchange the data presented that 4 01 02 12 22 31 32 the leader is not satisfied from this solution. Consequently, g5  x01  x02  x11  4x21  5x22  x32  20, application of the suggested method is discussed with a g6  4x01  3x02  2x11  x12  x21  x22  36, numerical model and the effectiveness of the solutions g7  5x01  2x02  x12  3x22  4x32  30, obtained by the suggested method is verified. Moreover, g8  2x01  x02  5x11  x21  2x32  20, from table 2, our suggested approach gives a more efficient g  x  x  2x  x  4x  x  8, solution comparing to the approaches of Ahlatcioglu and 9 01 02 11 22 31 32 Tiryaki [13]. g10  x01  3x11  2x12  x21  4x31  2x32  48, g  3x  5x  2x  5x Hence, our suggested algorithm can be easily extended 11 01 11 21 22  2x32 15, g  x  2x  x  x  3x  5x  x both the lower level and upper level with multiple objectives 12 01 02 11 12 21 31 32  60, (for example, [31]). x01, x02 , x11, x12 , x21, x22 , x31, x32  0 Step 7: There does not exists feasible solution for the above Acknowledgements problem with  1and   0.9. So the leader adjusts the 0 1 minimal acceptable level  1 by reducing  1 to 0 0 The author sincerely thank the anonymous reviewers and 0  0.9. editor-in-chief for their careful reading, constructive The above problem is solved with   0.9 and comments, and fruitful suggestions. 0   0.9using the Maple 18.02 software program, the optimal 1 The author confirms that there is no conflict of interest solutions for the above problem are regarding a financial supporter. x  2.640, x  0., x  0., x  2.008, 01 02 11 12 x  0.260, x  0.,x 1.340,x  0. 21 22 31 32 For this optimal solution, membership functions values are   f0   0.959,   f1   0.873,   f2   0.611,   f3   0.753 and objective values are f0  2.150, f1  0.668, f2  0.207, f3  4.340. 227 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 Table 2: Comparison of results of Example. f f f f   f0    f1    f2    f 3  d x 0 1 2 3  D The proposed method 2.150 0.668 0.207 4.340 0.959 0.873 0.611 0.753 0.919 0.934 0.480 Method in [13] 2.082 0.655 0.510 4.22 0.950 0.856 0.601 0.709  0.738 0.517 [14] Shih HS, Lai YJ & Lee ES (1996). Fuzzy approach References for multi-level programming problems. Computer & Operations Research, 23, 73–91. [1] Bracken J & McGill JM. (1973). Mathematical programs [15] Shih HS & Lee ES (2000). Compensatory fuzzy with optimization problems in the constraints. multiple level decision making. Fuzzy Sets and Systems, Operational Research, 21, 37–44. 114(1), 71-87. [2] Bracken J & McGill JM. (1974). Defense applications of [16] Sakawa M, Nishizaki I & Uemura Y (2000). mathematical programs with optimization problem in the Interactive fuzzy programming for two-level linear constraints. Operations Research, 22,1086–1096. fractional programming problems with fuzzy parameters. [3] Aiyoshi E & Shimizu K (1981). Hierarchical Fuzzy Sets and Systems, 115(1), 93-103. decentralized systems and its new solution by a barrier [17] Sakawa M & Nishizaki I (2001). Interactive fuzzy method. IEEE Transactions on Systems, Man, and programming for two-level linear fractional programming Cybernetics,11(6), 444–449. problems. Fuzzy Sets and Systems, 119(1), 31-40. [4] Anandalingam G & Apprey V (1991). Multi-level [18] Sakawa M & Nishizaki I (2002). Interactive fuzzy programming and conflict resolution. European Journal programming for two-level nonconvex programming of Operational Research, 51, 233–247. problems with fuzzy parameters through genetic [5] Ben-Ayed O & Blair CE (1990). Computational algorithms. Fuzzy Sets and Systems, 127(2), 185-197. difficulties of bilevel linear programming. Operations [19] Sakawa M, & Nishizaki I (2002). Interactive fuzzy Research, 38, 556–560. programming for decentralized two-level linear [6] Candler, W & Towersley R (1982). A linear two-level programming problems. Fuzzy Sets and Systems, 125(3), programming problem. Computers and Operations 301-315. Research, 9, 59–76. [20] Tiryaki F (2006). Interactive compensatory fuzzy [7] Bialas WF & Karwan MH (1984). Two-level linear programming for decentralized multi-level linear programming. Management Science, 30, 1004–1020. programming (DMLLP) problems. Fuzzy sets and systems, 157(23), 3072-3090. [8] Bard J& Falk J (1982). An explicit solution to the multi- level programming problem. Computers and Operations [21] Baky IA (2010). Solving multi-level multi-objective Research, 9, 77–100. linear programming problems through fuzzy goal programming approach. Applied Mathematical [9] Savard G & Gauvin J (1994). The steepest descent Modelling, 34(9), 2377-2387. direction for the nonlinear bilevel programming problem. Operations Research Letters, 15, 275–282. [22] Arora SR & Gupta R (2009). Interactive fuzzy goal programming approach for bilevel programming [10] Liu B (1998). Stackelberg–Nash equilibrium for problem. European Journal of Operational Research, multilevel programming with multiple followers using 194(2), 368-376. genetic algorithms. Computers & Mathematics with Applications, 36(7), 79–89. [23] Wang G, Wang X & Wan Z (2009). A fuzzy interactive decision making algorithm for bilevel multi- [11] Gao J & Liu B (2005). Fuzzy multilevel followers programming with partial shared variables programming with a hybrid intelligent algorithm. among followers. Expert Systems with Applications, Computers & Mathematics with Applications, 49(9-10), 36(7), 10471-10474. 1539–1548 [24] Yu PL (1973) A class of solutions for group [12] Lai YJ (1996). Hierarchical optimization: a decision problems. Management Science, 19(8), 936-946. satisfactory solution. Fuzzy Sets and Systems, 77(3), 321- 335. [25] Pal BB & Moitra BN (2003). A fuzzy goal programming procedure for solving quadratic bilevel [13] Ahlatcioglu M & Tiryaki F (2007). Interactive fuzzy programming problems. International Journal of programming for decentralized two-level linear fractional Intelligent Systems, 18(5), 529-540. programming (DTLLFP) problems. Omega, 35(4), 432- 450. 228 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Hasan Dalman, Vol.3, No.4, 2017 [26] Baky IA & Abo-Sinna MA (2013). TOPSIS for bi- Solve Decentralized Bi-level Multi-objective Fractional level MODM problems. Applied Mathematical Programming Problems. International Journal of Fuzzy Modelling, 37(3), 1004-1015. Systems, 1-10. [27] Sakawa M & Nishizaki I (2012). Interactive fuzzy [32] Singh, S., & Haldar, N. (2015). A New Method To programming for multi-level programming problems: a Solve Bi-Level Quadratic Linear Fractional Programming review. International Journal of Multicriteria Decision Problems. International Game Theory Review, 17(02), Making, 2(3), 241-266. 1540017. [28] Zheng Y, Liu J & Wan Z (2014). Interactive fuzzy [33] Kassem, M. A. E. H. (1995). Interactive stability of decision making method for solving bilevel programming multiobjective nonlinear programming problems with problem. Applied Mathematical Modelling, 38(13), 3136- fuzzy parameters in the constraints. Fuzzy Sets and 3141. Systems, 73(2), 235-243. [29] Sakawa, M., & Matsui, T. (2013). Interactive fuzzy [34] Sakawa, M., & Nishizaki, I. (2009). Cooperative random two-level linear programming based on level sets and noncooperative multi-level programming (Vol. 48). and fractile criterion optimization. Information Sciences, Springer Science & Business Media. 238, 163-175. [35] Pramanik, S., & Roy, T. K. (2007). Fuzzy goal [30] Emam, O. E. (2013). Interactive approach to bi- programming approach to multilevel programming level integer multi-objective fractional programming problems. European Journal of Operational Research, problem. Applied Mathematics and Computation, 223, 176(2), 1151-1166. 17-24. [36] Biswas A & Pal BB (2005). Application of fuzzy [31] Toksarı, M. D., & Bilim, Y. (2015). Interactive goal programming technique to land use planning in Fuzzy Goal Programming Based on Jacobian Matrix to agricultural system. Omega, 33(5), 391-398. 229 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Sait Koçak, Vol.3, No.4, 2017 Innovative Design for a Ball Worm Gear Mechanism Sait Koçak*‡ *Department of Mechatronics Engineering, Faculty of Technology, Pamukkale University, 20700 Denizli, Turkey (skocak@pau.edu.tr) ‡ Corresponding Author; Sait Koçak, Department of Mechatronics Engineering, Faculty of Technology, Pamukkale University, 20700 Denizli, Turkey, Tel: +90 258 296 4189, Fax: +90 258 296 41 96, skocak@pau.edu.tr Received: 31.10.2017 Accepted: 13.12.2017 Abstract- In this study, a new mechanism was designed and presented as an alternative to the worm gear mechanism without its disadvantages. This novel mechanism is referred to as a “ball worm gear mechanism”. The force of the worm gear mechanism is transmitted by a sliding movement, which leads to high operating temperature and low efficiency as well as wear on the bronze worm wheel. Therefore, a new mechanism was designed with balls placed into the helical grooves on the worm shaft which move by rolling on the worm wheel. The worm wheel of the newly designed mechanism is made of case-hardened steel, which is less expensive than bronze. Thus, it is anticipated that not only the cost of the worm wheel will be reduced, but also its life will be extended. Greater efficiency and elimination of the high operating temperature are expected to result from the power transmission achieved by the rolling motion of the new mechanism. Modelling of the design was carried out with SolidWorks software and the feasibility of the mechanism was confirmed. Keywords Worm gear mechanism, ball-screw mechanism, CAD, efficiency. 1. Introduction Worm gear mechanisms have very high conversion rates The invention of the wheel (4000-3500 BC) made it compared to other gear mechanisms. The conversion rates possible to transfer rolling motion to linear movement and laid achieved in several stages with reducers made up of other gear the foundations of the gear wheel system [1]. mechanisms can be attained in one step by means of the worm gear mechanism, thus allowing for the designing of lighter and Induced gear units of electric motors or internal cheaper constructions requiring less space [5, 6]. The screw combustion engines that are designed to reduce high rotational profile in the worm gear mechanism is usually dependent on speeds to the speeds required for the machines are called the manufacturing process. The five most commonly used reducers [2]. Gearboxes are specified according to the endless screws are the A, C, I, K, and N types [7]. The basic gearwheel which is used in the gear reducers, e.g., spur gear, concepts needed for calculating the dimensions of the worm helical gear, bevel gear and worm gear reducers. gear and worm wheel are shown in Figure 1 [8]. The present-day size of the gearbox is small when In addition to the abovementioned advantages of the worm compared that of the past. At 12 horsepower, the axis-to-axis gear mechanism, there are also disadvantages, such as low distance of a worm gear unit with a conversion ratio of 35 was efficiency, high operating temperatures and rapid worm wheel 356 mm in the year 1903, but now it is around 100 mm [3]. wear [9]. Moreover, due to adhesion wear when using the same material for both components, the worm gear and the One of the first screw profiles was invented by Archimedes worm wheel must be fabricated from two different materials in the 3rd century BC. Archimedes was able to transport water that have a low wear coefficient and are suitable for sliding. A upwards, thanks to the screw known by his name – the steel-bronze material pair is preferred at the medium-high Archimedes, or Archimedean, screw [4]. speed [6]. As bronze is expensive, it makes the cost of the endless screw mechanism high. 230 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Sait Koçak, Vol.3, No.4, 2017 screw and the nut in the ball screw is achieved by the balls. A sectional view of the ball screw is given in Figure 2 [22]. Friction line Fig. 1. Worm gear and worm wheel base Table 1. Efficiency of gear mechanisms Mechanism Efficiency (%) Fig. 2. Ball screw mechanism section Cylindrical gear mechanism 0.96 – 0.98 The balls in the ball screw roll away along the helical Bevel gear mechanism 0.95 – 0.97 grooves opened on the shaft. Thus, the sliding movement between the screw and the nut in the transmission shaft is Worm gear locking (γm > ρ) 0.6 – 0.8 converted into a rolling motion in the ball screw resulting in lower friction coefficients because rolling friction requires Worm gear anti-locking (γm ≤ ρ) 0.25 – 0.4 much less force than sliding friction. The difference between rolling and sliding friction is illustrated in Figure 3 [22]. The efficiency differences in various gear mechanisms are given in Table 1. The efficiency of the other gear mechanisms is around 95-98%, while for the worm gear mechanism type without locking it is as low as 25% [5]. Worm gear mechanisms are currently used in lifts and elevators, crane rope drums, textile machines, automobile and ship steering mechanisms, conveyors (banded, grilled and Fig. 3. Difference between sliding friction and rolling helical) and machine tools [3,10]. friction The worm gear mechanism has disadvantages which In the designing of the ball worm gear mechanism, the include low efficiency, rapid wear, overheating during concept of the rolling motion of balls in a nest on the shaft and operation and high cost. However, it is the advantage of its the rolling motion on the worm wheel was addressed. In order very wide field of application that led to the idea of a ball to make comparisons during the designing of the worm gear worm gear mechanism. mechanism, commercially available measurements were taken into consideration. To this purpose, a worm gear In this study, a new mechanism was designed to mechanism with an axis distance (e) of 80 mm and a cycle overcome the disadvantages of the worm gear mechanism. ratio (i) of 30 was taken as a reference. The newly designed mechanism was given the name of “ball worm gear mechanism”. The design of the mechanism was 2.1. Ball screw shaft design implemented using the SolidWorks package program and the mechanics were tested via the SolidWorks Motion Study In the design of the ball screw shaft, hemispherical module. cavities with a diameter (dw) of 8 mm were formed on a 2. Design of the ball worm gear mechanism helical 13 mm pitch (pa). In the design phase it was noted that the distance between the axes (e) was 80 mm and the cycle ratio (i) was 30. A technical drawing of the ball screw shaft is The design of gear mechanisms is a research area that has attracted many scientists and engineers over a number of years given in Figure 4. [11-21]. The use of cementation steel was planned for the production of the ball screw shaft and in the manufacturing The design of the ball worm gear mechanism was inspired stage, a lathe and vertical milling machine would be used. by the ball screw. The transmission of the force between the Shaft diameters were to be made on the lathe. To open the 231 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Sait Koçak, Vol.3, No.4, 2017 grooves on the shaft, a spherical end-mill cutter would be used wheel, firstly the material was to be brought to the desired in a vertical milling machine. length and diameter measurements on the lathe, and then the teeth would be formed with a specially designed milling cutter (Fig. 6). Fig. 6. Special semi-circular mouthed milling cutter Fig. 4. Technical drawing of ball screw shaft Commercially available milling cutters cannot be used in 2.2. Ball worm wheel design the manufacture of the ball worm wheel as they are module- mouthed. For this reason, a milling cutter with a semi-circular In the design of the ball worm wheel, semi-circular rim similar to that shown in Figure 6 was selected. profiles were used instead of module teeth. These profiles, through which the balls will roll, will be slightly larger than 2.3. Guide sleeve design the ball diameter to allow the mechanism to work more smoothly. Relevant information about the size of the opening The guide sleeve in the ball worm gear mechanism was will be provided by future experimental work. A technical designed to guide the balls within the ball screw mechanism drawing of the ball worm wheel is shown in Figure 5. like a ball nut and prevent the balls from spreading around during operation. A hemispherical gap was opened on the guide sleeve to allow for the working zones of the ball screw shaft and ball worm wheel equivalent gear. At the same time, the guide sleeve was fixed in the frame to accommodate axial reaction forces. A technical drawing of the guide sleeve is given in Figure 7. Fig. 5. Technical drawing of ball worm wheel Cementation steel was planned for use in the manufacture Fig. 7. Technical drawing of the guide sleeve of the ball worm wheel. In the production of the ball worm 232 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Sait Koçak, Vol.3, No.4, 2017 Conventionally produced steel was used as guide sleeve The ball worm wheel was designed as two parts for easy material. During the manufacturing phase, first, the inside and assembly and disassembly. The operation of the mechanics outside of the steel tube diameter will be machined on the was tested through the SolidWorks design program. lathe, then the hemispherical spaces will be formed using an end milling cutter on a machining center. An alternative In the Motion Study Module of the SolidWorks program, technique would be to manufacture the part by the casting it was observed that the mechanism worked smoothly when method and machining the necessary features on the side. contact was established between all parts and the rotational moment was provided via the input shaft. 2.4. Body design 2.6. Sizing of the ball worm gear mechanism For the ball worm gear mechanism, in place of a new body design, a commercially available worm gear gearbox body The pitch / pi connection was used instead of the module with an 80 mm distance between axes and a conversion ratio to size the ball worm gear mechanism. of 30 was used. The ball screw shaft clearance of the ready- made body was designed so that the guide sleeve had a smooth The pitch diameter of the worm wheel was calculated using fit. In addition, the radial screw holes were drilled on the body equation (1). to secure the guide sleeve axially. Current worm gear p mechanisms have cooling fins on the gearbox body because dm2 = ma . z2 = a . z2 (1) the operating temperatures are very high. This requirement is π expected to be omitted in the newly designed mechanism. 2.5. Assembly of the ball worm gear mechanism The shaft distance was calculated using equation (2). In the installation of the ball screw mechanism, the guide dm1+ dm2 is first placed in the housing and fixed axially and radially. e = (2) 2 The ball worm screw shaft, on which the ball bits are attached, is then inserted into the guide sleeve. In practice, the holding of the balls onto the shaft will be carried out with the aid of Givens: high lubricity. Deep-grooved ball bearings are installed on both sides of the assembled ball worm screw shaft. The Pitch, p = 13 mm assembly of the ball worm wheel is carried out in the same a way. A cross-section view of the assembled ball endless screw mechanism is given in Figure 8. Shaft distance, e = 80 mm Number of teeth, z1 = 1, z2 = 30 Ball diameter, dw = 8 mm Sizes: Pitch diameter of the ball worm wheel from equation (1) dm2 = 124.14 mm Pitch diameter of the ball worm shaft from equation (2) dm1 = 35.86 mm 3. Conclusion In the newly designed ball worm gear mechanism, the ball pieces slide on the worm screw shaft and proceed to move by rolling on the ball worm wheel. Since rolling friction requires less force than sliding friction, the ball worm gear mechanism will have higher efficiency than the conventional worm gear Fig. 8. Cross-section view of the assembled ball worm gear mechanism. mechanism In the ball worm gear mechanism, as the force transmission between the ball worm shaft and the ball worm 233 INTERNATIONAL JOURNAL of ENGINEERING TECHNOLOGIES-IJET Sait Koçak, Vol.3, No.4, 2017 wheel is effected by rolling, the mechanism will not generate [10] M. Pak, Sonsuz Vidalı Redüktörlerin Bilgisayar very high temperatures during operation. This situation will Yardımıyla Parametrik Tasarımı, Sakarya Üniversitesi Fen allow the gearbox oil to retain its long-term properties and bilimler Enstitüsü, Sakarya, 1998. make the mechanism more durable. [11] J.R. Colbourne, The use of oversize hobs to cut worm In the worm gear mechanism, since the force is gears, American Gear Manufacturers Association, 1989, pp transmitted by sliding, the worm wheel is usually produced 89. from bronze, which is a softer material than that used for the worm shaft. Bronze is expensive and exhibits rapid wear. In [12] M. De Donno, F.L. Litvin, Computerized design and the newly designed ball worm gear mechanism, the ball worm generation of worm gear drives with stable bearing contact wheel will be manufactured from cementation steel, which is and low transmission errors, ASME Journal of Mechanical much cheaper than bronze. This will both reduce the cost of Design, vol.121, No. 4, pp. 573–578, 1999. the mechanism and extend its life. [13] D.B. Dooner, A.A. Seirig, The Kinematic Geometry of Acknowledgments Gearing: a concurrent engineering approach, 3rd ed. John Wiley and Sons, New York, 1995. This study was performed in the framework of The Scientific Research Projects Unit of Pamukkale University, [14] H. S. Fang, C. B. Tsay, Mathematical model and bearing Project No: 2013FBE034. The author gratefully contacts of the zn-type worm gear set cut by oversize hob acknowledges the support of this institution. cutters, Mechanism and Machine Theory vol. 35, pp. 1689– 1708, 2000. References [15] S.A. Lagutin, Synthesis and application of general type [1] B. Salman, Sonsuz Vidalar ve Sonsuz Vida Karşılık Dişlisi worm gears with localized contact, Tech. Univ., Rektor, 1998. Helisel Dişli Matematik Modellemesi, İstanbul Teknik Üniversitesi Fen bilimler Enstitüsü, İstanbul, 2009. [16] F.L. Litvin, Development of gear technology and theory of gearing, NASA Reference Publication pp.1406-1500, 1998. [2] J. J. Uicker, G. R. Pennock, J. E. Shigley, Theory of Machines and Mechanisms, 3rd ed., Oxford University Press, [17] I.H. Seol, Design, generation and simulation of meshing New York, 2003. of worm-gear drive with longitudinally localized contacts, ASME Journal of Mechanical Design vol.122, no.2, pp. 201– [3] P. W. Crosher, Design and Application of the Worm Gear, 206, 2000. ASME Press, New York, 2002. [18] V.V. Simon, Computer aided loaded tooth contact [4] P. J. Kantert, Manual for Archimedean Screw Pump, analysis in cylindrical worm gears, Journal of Mechanical Hirthammer Verlag, 2008. Design vol.127, pp.973 – 981, 2005. [5] T. Şekercioğlu, Makine Elemanları Hesap Şekillendirme, [19] V.V. Simon, Influence of tooth errors and shaft 1st ed. Birsen Yayınevi, 2013, pp.327. misalignments on loaded tooth contact in cylindrical worm gears, Mechanism and Machine Theory vol.41, pp.707–724, [6] F.C. Babalık, Makine Elemanları ve Konstrüksiyon 2006. Örnekleri, 3rd ed., Nobel Yayınevi, 2008. [20] D.P. Townsend, Dudley’s Gear Handbook, 2nd ed., [7] ISO/TR 10828-1997, Worm Gears – Geometry of worm McGraw-Hill Inc., New York, 1992. profiles, 1997. [21] C. Zanzi, J.I. Pedrero, Application of modified geometry [8] D. Muhns, H. Wittel, D. Jannasch, J. Vobiek, Rolof / of face gear drive, Computer Methods in Applied Mechanics Matek Maschinenelemente, 19th, Vieweg Verlag, Wiesbaden, and Engineering, vol. 194, pp. 3047–3066, 2005. 2011. [22] Ball Screw. http://www.nsk.com/products/precision [9] S. Koçak, Bilyeli Sonsuz Vida Mekanizması Tasarımı ve machine/ballscrew/index.html #tab2 (accessed 07.08.17). İmalatı, Pamukkale Üniversitesi Fen bilimler Enstitüsü, Denizli, 2014. 234 INTERNATIONAL JOURNAL OF ENGINEERING TECHNOLOGIES-IJET Guide for Authors The International Journal of Engineering Technologies (IJET) seeks to promote and disseminate knowledge of the various topics of engineering technologies. The journal aims to present to the international community important results of work in the fields of engineering such as imagining, researching, planning, creating, testing, improving, implementing, using and asking. The journal also aims to help researchers, scientists, manufacturers, institutions, world agencies, societies, etc. to keep up with new developments in theory and applications and to provide alternative engineering solutions to current. The International Journal of Engineering Technologies is a quarterly published journal and operates an online submission and peer review system allowing authors to submit articles online and track their progress via its web interface. The journal aims for a publication speed of 60 days from submission until final publication. The coverage of IJET includes the following engineering areas, but not limited to: All filed of engineering such as; Chemical engineering  Biomolecular engineering  Materials engineering  Molecular engineering  Process engineering Civil engineering  Environmental engineering  Geotechnical engineering  Structural engineering  Transport engineering  Water resources engineering Electrical engineering  Computer engineering  Electronic engineering  Optical engineering  Power engineering Mechanical engineering  Acoustical engineering  Manufacturing engineering  Thermal engineering  Vehicle engineering Systems (interdisciplinary) engineering  Aerospace engineering  Agricultural engineering  Applied engineering  Biological engineering  Building services engineering  Energy engineering  Railway engineering  Industrial engineering  Mechatronics  Military engineering  Nano engineering  Nuclear engineering  Petroleum engineering Types of Articles submitted should be original research papers, not previously published, in one of the following categories, – Applicational and design studies. – Technology development, – Comparative case studies. – Reviews of special topics. – Reviews of work in progress and facilities development. – Survey articles. – Guest editorials for special issues. Editor-in-Chief and Associate Editors Editor-in-Chief: Prof. Dr. Mustafa BAYRAM Associate Editors: Prof. Dr. A. Burak POLAT Assoc. Prof. Dr. Baris SEVIM Asst. Prof. Dr. Ahmet AKTAS Asst. Prof. Dr. Yalcin CEKIC Asst. Prof. Dr. Ali ETEMADI Ethic Responsibilities The publication of an article in peer-reviewed “International Journal of Engineering Technologies” is an essential building block in the development of a coherent and respected network of knowledge. It is a direct reflection of the quality of the work. Peer-reviewed articles support and embody the scientific method. It is therefore important to agree upon standards of expected ethical behavior for all parties involved in the act of publishing: the author, the journal editor, the peer reviewer, the publisher and the society of society-owned or sponsored journals. All authors are requested to disclose any actual or potential conflict of interest including any financial, personal or other relationships with other people or organizations within three years of beginning the submitted work that could inappropriately influence, or be perceived to influence, their work. Submission of an article implies that the work described has not been published previously that it is not under consideration for publication elsewhere. The submission should be approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out, and that, if accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder. Upon acceptance of an article, authors will be asked to complete a “Copyright Form”. Acceptance of the agreement will ensure the widest possible dissemination of information. An e-mail will be sent to the corresponding author confirming receipt of the manuscript together with a “Copyright Form” form or a link to the online version of this agreement. Author Rights As a journal author, you retain rights for a large number of author uses, including use by your employing institute or company. These rights are retained and permitted without the need to obtain specific permission from IJET. These include:  The right to make copies (print or electronic) of the journal article for your own personal use, including for your own classroom teaching use;  The right to make copies and distribute copies (including via e-mail) of the journal article to research colleagues, for personal use by such colleagues for scholarly purposes;  The right to post a pre-print version of the journal article on internet web sites including electronic pre- print servers, and to retain indefinitely such version on such servers or sites for scholarly purposes  the right to post a revised personal version of the text of the final journal article on your personal or institutional web site or server for scholarly purposes  The right to use the journal article or any part thereof in a printed compilation of your works, such as collected writings or lecture notes. Article Style Authors must strictly follow the guide for authors, or their articles may be rejected without review. Editors reserve the right to adjust the style to certain standards of uniformity. Follow Title, Authors, Affiliations, Abstract, Keywords, Introduction, Materials and Methods, Theory/Calculation, Conclusions, Acknowledgements, References order when typing articles. The corresponding author should be identified with an asterisk and footnote. Collate acknowledgements in a separate section at the end of the article and do not include them on the title page, as a footnote to the title or otherwise. Abstract and Keywords: Enter an abstract of up to 250 words for all articles. This is a concise summary of the whole paper, not just the conclusions, and is understandable without reference to the rest of the paper. It should contain no citation to other published work. Include up to six keywords that describe your paper for indexing purposes. Abbreviations and Acronyms: Define abbreviations and acronyms the first time they are used in the text, even if they have been defined in the abstract. Abbreviations such as IEEE, SI, MKS, CGS, sc, dc, and rms do not have to be defined. Do not use abbreviations in the title unless they are unavoidable. Text Layout for Peer Review: Use single column layout, double spacing and wide (3 cm) margins on white paper at the peer review stage. Ensure that each new paragraph is clearly indicated. Present tables and figure legends in the text where they are related and cited. Number all pages consecutively; use 12 pt font size and standard fonts; Times New Roman, Helvetica, or Courier is preferred. Research Papers should not exceed 12 printed pages in two-column publishing format, including figures and tables. Technical Notes and Letters should not exceed 2,000 words. Reviews should not exceed 20 printed pages in two-column publishing format, including figures and tables. Equations: Number equations consecutively with equation numbers in parentheses flush with the right margin, as in (1). To make equations more compact, you may use the solidus ( / ), the exp function, or appropriate exponents. Italicize Roman symbols for quantities and variables, but not Greek symbols. Use an dash (–) rather than a hyphen for a minus sign. Use parentheses to avoid ambiguities in denominators. Punctuate equations with commas or periods when they are part of a sentence, as in C = a + b (1) Symbols in your equation should be defined before the equation appears or immediately following. Use “Eq. (1)” or “equation (1),” while citing. Figures and Tables: All illustrations must be supplied at the correct resolution: * Black and white and colour photos - 300 dpi * Graphs, drawings, etc - 800 dpi preferred; 600 dpi minimum * Combinations of photos and drawings (black and white and color) - 500 dpi In addition to using figures in the text, upload each figure as a separate file in either .tiff or .eps format during submission, with the figure number. Table captions should be written in the same format as figure captions; for example, “Table 1. Appearance styles.”. Tables should be referenced in the text unabbreviated as “Table 1.” References: Please ensure that every reference cited in the text is also present in the reference list (and viceversa). Any references cited in the abstract must be given in full. Unpublished results and personal communications are not recommended in the reference list, but may be mentioned in the text. Citation of a reference as “in press” implies that the item has been accepted for publication. Number citations consecutively in square brackets [1]. Punctuation follows the bracket [2]. Refer simply to the reference number, as in [3]. Use “Ref. [3]” or Reference [3]” at the beginning of a sentence: “Reference [3] was …”. Give all authors’ names; use “et al.” if there are six authors or more. For papers published in translated journals, first give the English citation, then the original foreign-language citation. Books [1] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol. 2. Oxford:Clarendon Press, 1892, pp.68-73. Journals [2] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron spectroscopy studies on magneto-optical media and plastic substrate interface”, IEEE Transl. J. Magn. Japan, vol. 2, pp. 740-741, August 1987. Conferences [3] Çolak I., Kabalci E., Bayindir R., and Sagiroglu S, “The design and analysis of a 5-level cascaded voltage source inverter with low THD”, 2nd PowerEng Conference, Lisbon, pp. 575-580, 18-20 March 2009. Reports [4] IEEE Standard 519-1992, Recommended practices and requirements for harmonic control in electrical power systems, The Institute of Electrical and Electronics Engineers, 1993. Text Layout for Accepted Papers: A4 page margins should be margins: top = 24 mm, bottom = 24 mm, side = 15 mm. Main text should be given in two column. The column width is 87mm (3.425 in). The space between the two columns is 6 mm (0.236 in). Paragraph indentation is 3.5 mm (0.137 in). Follow the type sizes specified in Table. Position figures and tables at the tops and bottoms of columns. Avoid placing them in the middle of columns. Large figures and tables may span across both columns. Figure captions should be centred below the figures; table captions should be centred above. Avoid placing figures and tables before their first mention in the text. Use the abbreviation “Fig. 1,” even at the beginning of a sentence. Type size Appearance (pts.) Regular Bold Italic 10 Authors’ affiliations, Section titles, Abstract references, tables, table names, first letters in table captions,figure captions, footnotes, text subscripts, and superscripts 12 Main text, equations, Authors’ names, a Subheading (1.1.) 24 Paper title Submission checklist: It is hoped that this list will be useful during the final checking of an article prior to sending it to the journal's Editor for review. Please consult this Guide for Authors for further details of any item. Ensure that the following items are present:  One Author designated as corresponding Author: • E-mail address • Full postal address • Telephone and fax numbers  All necessary files have been uploaded • Keywords: a minimum of 4 • All figure captions (supplied in a separate document) • All tables (including title, description, footnotes, supplied in a separate document)  Further considerations • Manuscript has been "spellchecked" and "grammar-checked" • References are in the correct format for this journal • All references mentioned in the Reference list are cited in the text, and vice versa • Permission has been obtained for use of copyrighted material from other sources (including the Web) • Color figures are clearly marked as being intended for color reproduction on the Web (free of charge) and in print or to be reproduced in color on the Web (free of charge) and in black-and-white in print. Article Template Containing Author Guidelines for Peer-Review First Author*, Second Author**‡, Third Author*** *Department of First Author, Faculty of First Author, Affiliation of First Author, Postal address **Department of Second Author, Faculty of First Author, Affiliation of First Author, Postal address ***Department of Third Author, Faculty of First Author, Affiliation of First Author, Postal address (First Author Mail Address, Second Author Mail Address, Third Author Mail Address) ‡ Corresponding Author; Second Author, Postal address, Tel: +90 312 123 4567, Fax: +90 312 123 4567,corresponding@affl.edu Received: xx.xx.xxxx Accepted:xx.xx.xxxx Abstract- Enter an abstract of up to 250 words for all articles. This is a concise summary of the whole paper, not just the conclusions, and is understandable without reference to the rest of the paper. It should contain no citation to other published work. Include up to six keywords that describe your paper for indexing purposes. Define abbreviations and acronyms the first time they are used in the text, even if they have been defined in the abstract. Abbreviations such as IEEE, SI, MKS, CGS, sc, dc, and rms do not have to be defined. Do not use abbreviations in the title unless they are unavoidable. Keywords- Keyword1; keyword2; keyword3; keyword4; keyword5. 2. Introduction Authors should any word processing software that is capable to make corrections on misspelled words and grammar structure according to American or Native English. Authors may get help by from word processor by making appeared the paragraph marks and other hidden formatting symbols. This sample article is prepared to assist authors preparing their articles to IJET. Indent level of paragraphs should be 0.63 cm (0.24 in) in the text of article. Use single column layout, double-spacing and wide (3 cm) margins on white paper at the peer review stage. Ensure that each new paragraph is clearly indicated. Present tables and figure legends in the text where they are related and cited. Number all pages consecutively; use 12 pt font size and standard fonts; Times New Roman, Helvetica, or Courier is preferred. Indicate references by number(s) in square brackets in line with the text. The actual authors can be referred to, but the reference number(s) must always be given. Example: "..... as demonstrated [3, 6]. Barnaby and Jones [8] obtained a different result ...." IJET accepts submissions in three styles that are defined as Research Papers, Technical Notes and Letter, and Review paper. The requirements of paper are as listed below:  Research Papers should not exceed 12 printed pages in two-column publishing format, including figures and tables.  Technical Notes and Letters should not exceed 2,000 words.  Reviews should not exceed 20 printed pages in two-column publishing format, including figures and tables. Authors are requested write equations using either any mathematical equation object inserted to word processor or using independent equation software. Symbols in your equation should be defined before the equation appears or immediately following. Use “Eq. (1)” or “equation (1),” while citing. Number equations consecutively with equation numbers in parentheses flush with the right margin, as in Eq. (1). To make equations more compact, you may use the solidus ( / ), the exp function, or appropriate exponents. Italicize Roman symbols for quantities and variables, but not Greek symbols. Use an dash (–) rather than a hyphen for a minus sign. Use parentheses to avoid ambiguities in denominators. Punctuate equations with commas or periods when they are part of a sentence, as in C = a + b (1) Section titles should be written in bold style while sub section titles are italic. 3. Figures and Tables 3.1. Figure Properties All illustrations must be supplied at the correct resolution:  Black and white and colour photos - 300 dpi  Graphs, drawings, etc - 800 dpi preferred; 600 dpi minimum  Combinations of photos and drawings (black and white and colour) - 500 dpi In addition to using figures in the text, Authors are requested to upload each figure as a separate file in either .tiff or .eps format during submission, with the figure number as Fig.1., Fig.2a and so on. Figures are cited as “Fig.1” in sentences or as “Figure 1” at the beginning of sentence and paragraphs. Explanations related to figures should be given before figure. Figures and tables should be located at the top or bottom side of paper as done in accepted article format. Figure 1. Engineering technologies. Table captions should be written in the same format as figure captions; for example, “Table 1. Appearance styles.”. Tables should be referenced in the text unabbreviated as “Table 1.” Table 1. Appearance properties of accepted manuscripts Type size Appearance (pts.) Regular Bold Italic 10 Authors’ affiliations, Abstract, keywords, Abstract references, tables, table names, figure captions, footnotes, text subscripts, and superscripts 12 Main text, equations, Authors’ names, Subheading Section titles (1.1.) 24 Paper title 4. Submission Process The International Journal of Engineering Technologies operates an online submission and peer review system that allows authors to submit articles online and track their progress via a web interface. Articles that are prepared referring to this template should be controlled according to submission checklist given in “Guide f Authors”. Editor handles submitted articles to IJET primarily in order to control in terms of compatibility to aims and scope of Journal. Articles passed this control are checked for grammatical and template structures. If article passes this control too, then reviewers are assigned to article and Editor gives a reference number to paper. Authors registered to online submission system can track all these phases. Editor also informs authors about processes of submitted article by e-mail. Each author may also apply to Editor via online submission system to review papers related to their study areas. Peer review is a critical element of publication, and one of the major cornerstones of the scientific process. Peer Review serves two key functions:  Acts as a filter: Ensures research is properly verified before being published  Improves the quality of the research 5. Conclusion The conclusion section should emphasize the main contribution of the article to literature. Authors may also explain why the work is important, what are the novelties or possible applications and extensions. Do not replicate the abstract or sentences given in main text as the conclusion. Acknowledgements Authors may acknowledge to any person, institution or department that supported to any part of study. References [1] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol. 2. Oxford:Clarendon Press, 1892, pp.68-73. (Book) [2] H. Poor, An Introduction to Signal Detection and Estimation, New York: Springer-Verlag, 1985, ch. 4. (Book Chapter) [3] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron spectroscopy studies on magneto-optical media and plastic substrate interface”, IEEE Transl. J. Magn. Japan, vol. 2, pp. 740-741, August 1987. (Article) [4] E. Kabalcı, E. Irmak, I. Çolak, “Design of an AC-DC-AC converter for wind turbines”, International Journal of Energy Research, Wiley Interscience, DOI: 10.1002/er.1770, Vol. 36, No. 2, pp. 169-175. (Article) [5] I. Çolak, E. Kabalci, R. Bayindir R., and S. Sagiroglu, “The design and analysis of a 5-level cascaded voltage source inverter with low THD”, 2nd PowerEng Conference, Lisbon, pp. 575-580, 18-20 March 2009. (Conference Paper) [6] IEEE Standard 519-1992, Recommended practices and requirements for harmonic control in electrical power systems, The Institute of Electrical and Electronics Engineers, 1993. (Standards and Reports) Article Template Containing Author Guidelines for Accepted Papers First Author*, Second Author**‡, Third Author*** *Department of First Author, Faculty of First Author, Affiliation of First Author, Postal address **Department of Second Author, Faculty of First Author, Affiliation of First Author, Postal address ***Department of Third Author, Faculty of First Author, Affiliation of First Author, Postal address (First Author Mail Address, Second Author Mail Address, Third Author Mail Address) ‡ Corresponding Author; Second Author, Postal address, Tel: +90 312 123 4567, Fax: +90 312 123 4567,corresponding@affl.edu Received: xx.xx.xxxx Accepted:xx.xx.xxxx Abstract- Enter an abstract of up to 250 words for all articles. This is a concise summary of the whole paper, not just the conclusions, and is understandable without reference to the rest of the paper. It should contain no citation to other published work. Include up to six keywords that describe your paper for indexing purposes. Define abbreviations and acronyms the first time they are used in the text, even if they have been defined in the abstract. Abbreviations such as IEEE, SI, MKS, CGS, sc, dc, and rms do not have to be defined. Do not use abbreviations in the title unless they are unavoidable. Keywords Keyword1, keyword2, keyword3, keyword4, keyword5. 1. Introduction given. Example: "..... as demonstrated [3,6]. Barnaby and Jones [8] obtained a different result ...." Authors should any word processing software that is capable to make corrections on misspelled words and IJET accepts submissions in three styles that are defined grammar structure according to American or Native English. as Research Papers, Technical Notes and Letter, and Review Authors may get help by from word processor by making paper. The requirements of paper are as listed below: appeared the paragraph marks and other hidden formatting  Research Papers should not exceed 12 printed pages symbols. This sample article is prepared to assist authors in two-column publishing format, including figures and preparing their articles to IJET. tables. Indent level of paragraphs should be 0.63 cm (0.24 in) in  Technical Notes and Letters should not exceed the text of article. Use single column layout, double-spacing 2,000 words. and wide (3 cm) margins on white paper at the peer review stage. Ensure that each new paragraph is clearly indicated.  Reviews should not exceed 20 printed pages in two- Present tables and figure legends in the text where they are column publishing format, including figures and tables. related and cited. Number all pages consecutively; use 12 pt Authors are requested write equations using either any font size and standard fonts; Times New Roman, Helvetica, mathematical equation object inserted to word processor or or Courier is preferred. Indicate references by number(s) in using independent equation software. Symbols in your square brackets in line with the text. The actual authors can equation should be defined before the equation appears or be referred to, but the reference number(s) must always be immediately following. Use “Eq. (1)” or “equation (1),” while citing. Number equations consecutively with equation .tiff or .eps format during submission, with the figure number numbers in parentheses flush with the right margin, as in Eq. as Fig.1., Fig.2a and so on. Figures are cited as “Fig.1” in (1). To make equations more compact, you may use the sentences or as “Figure 1” at the beginning of sentence and solidus ( / ), the exp function, or appropriate exponents. paragraphs. Explanations related to figures should be given Italicize Roman symbols for quantities and variables, but not before figure. Greek symbols. Use an dash (-) rather than a hyphen for a minus sign. Use parentheses to avoid ambiguities in denominators. Punctuate equations with commas or periods when they are part of a sentence, as in C = a + b (1) Section titles should be written in bold style while sub section titles are italic. 6. Figures and Tables 6.1. Figure Properties All illustrations must be supplied at the correct resolution:  Black and white and colour photos - 300 dpi Fig. 1. Engineering technologies.  Graphs, drawings, etc - 800 dpi preferred; 600 dpi Figures and tables should be located at the top or bottom minimum side of paper as done in accepted article format. Table captions should be written in the same format as figure  Combinations of photos and drawings (black and captions; for example, “Table 1. Appearance styles.”. Tables white and colour) - 500 dpi should be referenced in the text unabbreviated as “Table 1.” In addition to using figures in the text, Authors are requested to upload each figure as a separate file in either Table 1. Appearance properties of accepted manuscripts Appearance Type size (pts.) Regular Bold Italic Main text, section titles, authors’ affiliations, abstract, 10 keywords, references, tables, table names, figure captions, Abstract- Subheading (1.1.) equations, footnotes, text subscripts, and superscripts 12 Authors’ names, 24 Paper title 6.2. Text Layout for Accepted Papers the text. Use the abbreviation “Fig. 1,” even at the beginning of a sentence. A4 page margins should be margins: top = 24 mm, bottom = 24 mm, side = 15 mm. The column width is 87mm 7. Submission Process (3.425 in). The space between the two columns is 6 mm (0.236 in). Paragraph indentation is 3.5 mm (0.137 in). The International Journal of Engineering Technologies Follow the type sizes specified in Table. Position figures and operates an online submission and peer review system that tables at the tops and bottoms of columns. Avoid placing allows authors to submit articles online and track their them in the middle of columns. Large figures and tables may progress via a web interface. Articles that are prepared span across both columns. Figure captions should be centred referring to this template should be controlled according to below the figures; table captions should be centred above. submission checklist given in “Guide f Authors”. Editor Avoid placing figures and tables before their first mention in handles submitted articles to IJET primarily in order to control in terms of compatibility to aims and scope of Journal. Articles passed this control are checked for grammatical and template structures. If article passes this control too, then reviewers are assigned to article and Editor gives a reference number to paper. Authors registered to online submission system can track all these phases. Editor also informs authors about processes of submitted article by e-mail. Each author may also apply to Editor via online submission system to review papers related to their study areas. Peer review is a critical element of publication, and one of the major cornerstones of the scientific process. Peer Review serves two key functions:  Acts as a filter: Ensures research is properly verified before being published  Improves the quality of the research 8. Conclusion The conclusion section should emphasize the main contribution of the article to literature. Authors may also explain why the work is important, what are the novelties or possible applications and extensions. Do not replicate the abstract or sentences given in main text as the conclusion. Acknowledgements Authors may acknowledge to any person, institution or department that supported to any part of study. References [7] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol. 2. Oxford:Clarendon Press, 1892, pp.68-73. (Book) [8] H. Poor, An Introduction to Signal Detection and Estimation, New York: Springer-Verlag, 1985, ch. 4. (Book Chapter) [9] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron spectroscopy studies on magneto-optical media and plastic substrate interface”, IEEE Transl. J. Magn. Japan, vol. 2, pp. 740-741, August 1987. (Article) [10] E. Kabalcı, E. Irmak, I. Çolak, “Design of an AC- DC-AC converter for wind turbines”, International Journal of Energy Research, Wiley Interscience, DOI: 10.1002/er.1770, Vol. 36, No. 2, pp. 169-175. (Article) [11] I. Çolak, E. Kabalci, R. Bayindir R., and S. Sagiroglu, “The design and analysis of a 5-level cascaded voltage source inverter with low THD”, 2nd PowerEng Conference, Lisbon, pp. 575-580, 18-20 March 2009. (Conference Paper) [12] IEEE Standard 519-1992, Recommended practices and requirements for harmonic control in electrical power systems, The Institute of Electrical and Electronics Engineers, 1993. (Standards and Reports) INTERNATIONAL JOURNAL OF ENGINEERING TECHNOLOGIES (IJET) COPYRIGHT AND CONSENT FORM This form is used for article accepted to be published by the IJET. Please read the form carefully and keep a copy for your files. TITLE OF ARTICLE (hereinafter, "The Article"): ………..…………………………………………………....................………………………………………………………… ………..…………………………………………………....................…………………………………………………………… …………..………………………………………………………………………..…………………………………………… LIST OF AUTHORS: ………..……………………………………………………………...…..……………………………………………………… ………..…………………………………………………....................…………………………………………………………… ……………………………………………………………………..……………………………………………..…………… CORRESPONDING AUTHOR’S (“The Author”) NAME, ADDRESS, INSTITUTE AND EMAIL: ………………………………………………………………………..………………………………………………………… ………..…………………………………………………....................…………………………………………………………… …………..………………………………………………………………………..…………………………………….…… COPYRIGHT TRANSFER The undersigned hereby transfers the copyright of the submitted article to International Journal of Engineering Technologies (the "IJET"). The Author declares that the contribution and work is original, and he/she is authorized by all authors and/or grant-funding agency to sign the copyright form. Author hereby assigns all including but not limited to the rights to publish, distribute, reprints, translates, electronic and published derivates in various arrangements or any other versions in full or abridged forms to IJET. IJET holds the copyright of Article in its own name. Author(s) retain all rights to use author copy in his/her educational activities, own websites, institutional and/or funder’s web sites by providing full citation to final version published in IJET. The full citation is provided including Authors list, title of the article, volume and issue number, and page number or using a link to the article in IJET web site. Author(s) have the right to transmit, print and share the first submitted copies with colleagues. Author(s) can use the final published article for his/her own professional positions, career or qualifications by citing to the IJET publication. Once the copyright form is signed, any changes about the author names or order of the authors listed above are not accepted by IJET. Authorized/Corresponding Author Date/ Signature