Algoritam za tehnološku klasifikaciju na temelju CAD 3D modela izratka

Antolić, Dražen (2017) Algoritam za tehnološku klasifikaciju na temelju CAD 3D modela izratka. = Algorithm for technological classification based on CAD 3D model of the workpiece. Doctoral thesis , Sveučilište u Zagrebu, Fakultet strojarstva i brodogradnje, UNSPECIFIED. Mentor: Ćosić, Predrag.

[img]
Preview
Text
Antolic _phd_2017.pdf - Published Version Jezik dokumenta:Croatian

Download (31MB) | Preview

Abstract (Croatian)

Rad se bavi istraživanjem mogućnosti automatizirane tehnološke klasifikacije na temelju CAD 3D modela izratka, što je nužan uvjet za automatizirano projektiranje osnovnih tehnoloških procesa, odnosno procjenu troškova izrade. Postavljena je hipoteza istraživanja: „Moguća je automatizirana tehnološka klasifikacija izradaka na temelju CAD 3D modela“. Pretpostavljeno je da u digitalnom zapisu CAD 3D modela postoje tehnološke značajke izratka koje je moguće izdvojiti u posebnu datoteku, temeljem koje se može definirati tehnološki tip izratka. Na temelju tehnološkog tipa izratka i izdvojenih značajki iz 3D modela, moguće je definirati i tehnološki postupak izrade. Istraživanje tehnoloških značajki izratka je izvršeno na CAD 3D modelima koji su kreirani u CAD softveru PTC/Creo 4.0. Tehnološki tip izratka definiran je kao kod sa 14 kodnih mjesta od kojih svako može poprimiti jednoznamenkastu cjelobrojnu vrijednost. Devet kodnih mjesta mogu poprimiti vrijednosti od 1 do 9, a pet kodnih mjesta može poprimiti vrijednosti od 0 do 9. Tehnološkim kodom izrade definira se 16 osnovnih elemenata tehnološkog procesa. Kodiranje se može izvršiti sa različitom detaljnošću, a ovim istraživanjem definirana je potreba za pojedinim elementima u tehnološkom procesu izrade. Binarno definiranih 16 osnovnih elemenata tehnološkog procesa moguće je opisati sa 6 kodnih mjesta. Svako kodno mjesto binarne klasifikacije može poprimiti cjelobrojnu vrijednost od 0 do 8. Tijekom istraživanja razvijena je posebna baza podataka i posebna računalna aplikacija za povezivanje novorazvijene baze podataka i CAD softvera PTC/Creo 4.0. Baza podataka nazvana je OTP baza podataka, a novorazvijena računalna aplikacija BTP aplikacija. Set tehnoloških značajki 3D modela, koji se automatizirano dohvaća pomoću BTP aplikacije i zapisuje u OTP bazu podataka predstavlja digitalni format zapisa tehnoloških značajki CAD 3D modela. Ovaj set podataka je u potpunosti standardiziran, te ga je moguće iz OTP baze podataka „izvoziti“ kao samostalni skup ulaznih podataka u neki od softvera za obradu podataka. Nazvan je btp zapis. Ovim istraživanjem napravljeni su algoritmi za obradu podataka u samoj OTP bazi podataka, što je rezultiralo egzaktnim kodiranjem tehnološkog tipa izratka. To znači da nema potrebe za uzorkom na temelju kojeg bi se matematičko statističkim metodama zaključivalo o pojedinim kodnim mjestima koda izratka. Ostvareno je rješenje kojim je moguće CAD 3D model svrstati u određeni tehnološki tip izratka samo na temelju značajki CAD 3D modela, ne ovisno o bilo kojim drugim vanjskim veličinama. Napravljeni algoritmi u OTP bazi podataka koji se temelje na tehnološkom kodu izratka, uz uključivanje i točno određene podatke iz seta dohvaćenih tehnoloških podataka rezultiraju i egzaktnom binarnom klasifikacijom izrade. Upis podataka i svi algoritmi u bazi podataka provode se potpuno automatizirano. Istraživanjem je postavljena hipoteza u cijelosti potvrđena. Uzimajući u obzir i ranija istraživanja [1]; [28] može se govoriti o novoj znanstvenoj metodi za projektiranje osnovnih tehnoloških procesa sa ciljem brze procjene cijene koštanja budućeg izratka na temelju njegova 3D modela. Metoda je nazvana OTP metoda, a njen daljnji razvoj ići će u detaljnost kodiranje izrade. Ostvarena binarna klasifikacija može se dalje razraditi uključivanjem više detalja o elementima osnovnog tehnološkog procesa. Razvila bi se dekadska i posebna tehnološka klasifikacija.

Abstract

The research into the possibility of estimating product development time based on its drawing (hard copy 2D drawing), completed in 2007, confirmed the possibility of determining development time by regression models, which is described in detail in the author's master's thesis. The research results, the equations correlating 2D drawing properties to the required development time, enable automated product manufacturing cost calculation. The solution also implies the possibility of automated quantification of the basic parameters of the technological process. This means that a technological process can be designed without the involvement of an engineering technologist. While easy quantification of 2D drawing features does not require the presence of a technologist for designing a technological process, it still requires some human input (less qualified persons – lower cost of labor), which means this is not fully automated technological process designing. Since a 2D drawing is nowadays only a graphical representation of the digital product file, research can be based on the properties of that very same digital file. Since a "digital" drawing is derived from a CAD 3D model of the future product, and given that 3D product modeling is no longer a mere trend, but an everyday practice, there arises a need to carry out research on the basis of CAD 3D product model properties, more so because of at least two basic advantages: 1. CAD 3D model digital file properties can be automatically transferred into a technological process design algorithm (with no need for the presence of an expert in the process). 2. Compared to a 2D drawing, a digital CAD 3D model rendering contains more properties that are potentially significant for defining the technological process. It should be also noted that the very application for technological process design can be integrated into a CAD 3D model creation software, which erases the line between product creation and technological process design. Product creation and technological process design become a unified whole. Given that one and the same product can be made in multiple different ways, and that the optimality of a technological process depends on criteria that may vary according to priorities (production cost, delivery deadline, available production resources...), it is possible for a single product to have multiple technological processes. Considering the requirements and design method, technological processes can be categorized into specific groups. Technological processes that are designed in an automated way (no technologist involved), to be implemented in an average (typical) manufacturing system are categorized into the group of basic technological processes (BTP), with the aim of determining the expected cost. For the purposes of this paper, the elements of a basic technological process have been defined in detail, as presented in a dedicated chapter. One of the goals of this research is to determine whether there is a possibility of automated retrieval of the technological properties of a CAD 3D model, based on which an algorithm for technological classification of products could be developed. This is also a prerequisite for the automation of technological process designing, i.e. the basis for the development of a new scientific method related to rapid cost assessment for a future product based on its CAD 3D model. The activities and results of this research, which fully validates the presented hypothesis, are described in several chapters of this paper. Chapter one, as an introduction provides an overview of the current status relating to the research in question. The author's previous studies are described as well, and a research plan is defined along with its accompanying flowchart. The earlier studies have confirmed the possibility of assessing production time via a system of regression equations for certain technological types of products. A problem which was not solved before this research using the researched method concerns the possibility of automated product classification into predetermined technological groups. In other words, the possibility of technological classification of products needs to be explored. The second chapter is dedicated to the technological process design method aimed at rapid product cost assessment. This is a potential new scientific method called the OTP method, i.e. basic technological process method (OTPm). Three basic elements of OTPm have been defined, specifically: OTP process, OTP system and OTP database. The OTP process comprises 6 basic activities, among which those relating to data retrieval from a CAD 3D model (3DM) are especially pertinent to this research, as well as the processing of such data in terms of product and manufacture coding. The OTP system necessary to enable the implementation of the OTP process comprises 8 core elements. Those are predetermined potential independent and dependent variables and the anticipated method constraints. The independent variables are related to 3DM properties, the dependent ones to the basic parameters of the technological process, and the constraints arise out of the current manufacturing potential or the normal state of manufacturing systems. A dedicated OTP database (OTPbp) has been, as part of this research, developed for the purposes of OTPm. Along with the hierarchical structure of the tables which can be used to store data related to 3DM properties, the OTPbp includes predetermined data on system elements (knowledge base) and algorithms for process activities implementation (artificial intelligence elements). MS Access database was used as a software solution. It can be stated that, given its limitations, MS Access was barely adequate for the purposes of this research. In future studies, OTPbp should by all means be rebuilt using more powerful software. The third chapter lays out the research hypothesis: "Automated technological classification of products based on CAD 3D models is possible." Five criteria on the basis which it can be precisely determined whether the research validates the hypothesis have been defined. The criteria imply the existence of a computer application for 3DM data retrieval. The fourth chapter mostly pertains to the 3DM data retrieval computer application. The existence or development of such application is also implied by the second criterion concerning the hypothesis. The research resulted in an application linking CAD software with OTPbp. It was named BTP application, or BTP, for short. The core functions of the newly developed application are two-fold. On the one hand, it serves for entering parameters into a 3DM, while on the other hand it serves for entering 3DM technological properties data into OTPbp. However, to achieve that, a series of algorithms have been built into BTP with the purpose of retrieving data from the 3DM. BTP analyzes the 3DM as a whole and derives technologically significant data from it, rendering them into a digital file format containing the technological properties of the 3DM. The extracted dataset has been fully defined and it is called ".btp", suggesting the intent of the author of this paper that in the future it should become one of the standard file formats in CAD applications. Such as, for instance, ".prt", ".stl", or ".pdf"... Chapter five describes data processing, going further into the details of the functions of the said BTP. Namely, in the development of BTP it was observed that such an extensive set of technological data could be retrieved that exact technological classification would be possible just on the basis on those data. Therefore, there would be no need for the mathematical and statistical processing of data from a sample that would be formed through the research. Algorithms can be integrated into BTP that not only enter parameters into a 3DM and retrieve available 3DM properties but also enable quantifying indirectly available 3DM properties. One option is to add a property into the 3DM tree and further analyze the 3DM by subsequent action from BTP upon parameters and added properties. An example is an additional property called "Hole", which makes it possible to determine the existence of a pass-through hole in the product. Developing this method further in BTP enables the pass-through hole's diameter to be quantified (if it exists), but also for the 3DM circumcircle's diameter to be measured. Another possibility is to use special tools provided by CAD software, based on which a 3DM can be researched. For this research, a tool was used that enables a beam to be projected onto a 3DM and thus determine the geometrical characteristics of the 3DM on the basis of its intersection with the 3DM planes. While the method is very simple and yields only the intersection coordinates information, the author of this research has devised several tools that can make the 3DM beam projection method a significant 3DM geometry data source. In the conducted research, only the compound tool called "Radial Beam Set" was used and it is described in detail in this chapter. However, the method itself and the BTP geometry analysis tools are described in more detail in the annexes to this paper, which points to numerous unexplored possibilities of use in technological classification. To name just a few examples of the tools: "Sieve", "Ring", "Disc" and "Comb". The measurement method can also be a combination of results obtained by using different tools. Alongside being processed in BTP, the data are processed in OTPbp after entry using the developed algorithms, which are integrated in the database itself. By analyzing the required data from the .btp set, a system of queries extracts the data that are directly required for defining dependent variables. Those dependent variables describe the technological type of product, i.e. the type of technological process. Two sets of data for the technological classification of a product are formed. The product coding dataset comprises 44 different data. The manufacture coding dataset comprises 39 data. Chapter six addresses product coding. In accordance with the product coding system as defined in chapter one, this chapter details the algorithms built into OTPbp. All of the algorithms for determining the value of each coding position of the OTP product code (OTPk). It is a coding system with 14 coding positions, 12 of which are categorized into 3 groups of 4 technological properties. The groups refer to product properties: "Shape", "Volume", "Complexity". The remaining two coding positions are "Product type" as the first digit and "Quantity" as the last digit of the OTPk code. Each coding place can be assigned a single-digit value. With respect to the code digit value, the same rule applies to each OTPk coding position: "The more technologically acceptable the property is, the lower the code number". It is understood that the more technologically acceptable property is the one which leads to a lower cost of manufacturing. Code value zero (0) means that the property does not even exits, so there is no expenditure arising from it. The technological features related to the specifics of the 3DM shape and heat treatment and surface protection requirements may or may not be represented in the 3DM, and they can be quantified by numbers from 0 to 9. The remaining 9 coding positions may be assigned values from 1 to 9. Chapter seven describes the procedure of manufacture coding, as the final activity in technological classification. With respect to the level of detail in elaborating the elements of the basic technological process, three instances of manufacture coding are possible. 1. Binary, which merely determines the need for an element in the technological process. 2. Decimal, where coding positions may be assigned two-digit values, which enables defining specific technological procedures as operations in the technological process. 3. A special, third instance is coding that would be carried out for a specific company, which would enable to specify the necessary technological procedures and the machine on which the technological operation would be carried out. One of the goals of this research is binary manufacture classification. Therefore, the paper describes the algorithms for quantifying all 16 elements of the basic technological process. All 16 binary-classified data can be represented by the 6 coding positions of OTPbk:  BK1-Primary process: Pre-shaping – Reshaping – Assembling  BK2-Secondary processes: Material removal – Heat – Surface  BK3-Starting material: Semi-finished products – Measurements – Mass  BK4-Input procedures: Cutting – Metalworking – Special procedures  BK5-Basic procedures: Turning – Drilling – Milling  BK6-Output procedures: Grinding – Additional procedures – Finishing procedures The data related to the secondary processes of heat treating and surface protection are additionally elaborated in such a way that not only the necessity of processing but also the exact technological action in the required processing is specified already in the binary classification. For the third coding position of OTPbk, concerning the starting material, the possibility of a more detailed elaboration is presented. This includes a description of the method by which the measurements and the required material mass for a measuring unit (typically kg/piece) of the product can be quantified on the basis of the specified shape of the starting material. In this manner, partial decimal manufacture classification is already achieved in the binary code. Chapter eight is an overview of the results achieved by the research. Since the paper is a dissertation, the emphasis is on validating the presented hypothesis, specifically with respect to the defined hypothesis validation criteria. Based on the criteria analysis, it is shown that the research has achieved a result that is even better than anticipated. Moreover, the results of this research and those of the author's previous studies, as a consequence of the new approach to the subject of research, could be the foundation for creating a new scientific method for engineering basic technological processes aimed at rapid 3DM-based product cost assessment. This is a consequence of additional elements to the solution that this research has led to, and at the same time a significant innovation compared to the present status. Those are the newly developed BTP, .btp, OTPbp and partial solutions within the CAD software itself. The said elements have already been developed through this research to a degree where they can be applied in practice. Not only does this paper demonstrate the possibility of automated 3DM-based technological classification, but a concrete, functional algorithm to confirm it has been created as part of the research. The research has fully validated the set hypothesis, which is a prerequisite for the realization of a commercial software. Therefore, further development of OTPm is to be expected. Given that it would cover a wide area, further development is possible through a series of multidisciplinary studies that could be realized through scientific, developmental, professional or even business projects. The research described in this paper is by all means the foundation for any future studies. Irrespective of which software would be used for developing a new database, its structure will be based upon the OTPbp structure described in this paper. BTP development, in addition to what is described in chapters 4 and 5 of the dissertation, will certainly be based on even more detailed descriptions provided in annexes [C]; [D]; [E]. Chapter nine, the conclusion of the research is provided only on a single page of the dissertation, which is why it is fully translated in this extended abstract into English, as described in the thesis: "The research validates the hypothesis regarding the possibility of automated technological classification of products on the basis of a CAD 3D product model." The hypothesis "CAD 3D model digital file also contains the data on the technological process of manufacturing“ has been validated given the fulfillment of all 5 of the criteria that were set for it. With slight limitations related to the modeling method in the PTC/Creo 4.0 software, the possibility of automated "retrieval" of numerous technological properties of CAD 3D models has been achieved. Properties that initially were not planned for, or it was not to be expected that they could be retrieved, have been extracted, which confirms that the research results are even better than expected. Extraction of properties from a 3D model into the database is enabled by using a BTP application, self-developed over the course of the research, which links the PTC/Creo 4.0 software to the OTP database, which was also self-developed over the course of this research. The result of the functioning of the BTP application is a digital file containing the technological properties of the CAD 3D product model. The set of technologically significant data is fully defined and recorded in the OTP database in the corresponding tables so that each registered information can be reached in order to be used for technological classification. The processing of the retrieved data in the OTP database creates the possibility to define the type of product and achieve precise coding of all 14 code digits. This means that there is no need for sampling and using statistical methods to quantify coding positions. This surely indicates that even in this segment a better result than expected has been achieved. For every 3DM, the manufacturing type, i.e. the technological process, can be defined on a binary level. This means that for every type of processing as a primary process, or every technological operation as an operational procedure, it is specified whether it is required in the technological process or not. This achievement is also a direct validation of the realization of automated technological classification. This result too has been realized through an exact method, only on the basis of the product code and the digital record of the technological properties of the CAD 3D model, which also points to a solution better than expected. Along with the validation of the hypothesis, the research has yielded a series of additional solution, most notably the BTP application, which has grown from the initial idea to link the CAD 3D model with the OTP database into an application with numerous functions. Most notable are the functions enabling the BTP to act upon CAD 3D model properties in order to retrieve more detailed geometrical data. While it is possible to develop additional classification systems, in terms of the solution's further development, numerous scientific, developmental and professional projects can be carried out; they would be implemented with the aim to build a commercial OTP software as an independent computer application. Overall, it can be claimed that the research has fully justified its purpose and can be used as basis for the development of a new scientific method for rapid manufacturing cost assessment – the OTP method.” Therefore, the paper researches the possibility of automated technological classification on the basis of a CAD 3D product model, which is an essential requirement for automated design of basic technological processes or manufacturing cost assessment. It is assumed that a CAD 3D model digital file contains the technological properties of a product that can be extracted into a separate file on the basis of which the technological type of product can be defined. On the basis of the technological type of product and the properties extracted from the 3D model, the technological procedure of manufacturing can be defined as well. The research on the technological properties of products was carried out on CAD 3D models created in PTC/Creo 4.0 CAD software. The technological type of product is defined as a code with 14 coding positions, where each position can be assigned a single-digit, whole-number value. Nine of the coding positions can be assigned the values from 1 to 9, and five coding positions can be assigned the values from 0 to 9. The technological code for manufacturing defines 16 basic elements of the technological process. Coding can be done at varying levels of details, and in this research the necessity for individual elements in the technological process of manufacturing was defined. The 16 basic elements of the technological process defined in a binary way can be described with 6 coding positions. Each coding position of the binary classification can be assigned a whole-number value from 0 to 8. During the research, a special database was developed as well as a dedicated computer application for linking the newly developed database to the PTC/Creo 4.0 software. The database was called the OTP database, and the new computer application was called the BTP application. The 3D model's technological properties set, which is auto-retrieved by the BTP application and registered in the OTP database represents the digital file format for CAD 3D model technological properties. This dataset is fully standardized and can be exported from the OTP database as a self-contained set of input data into a data processing software. It is called the btp file. As part of this research, algorithms for data processing in the OTP database itself were created, which has resulted in the exact coding of the technological type of product. This means that there is no need for a sample based on which individual coding positions of the product code would be extrapolated using mathematical and statistical methods. A solution was created by which a CAD 3D model can be classified as a certain technological type of product only on the basis of the CAD 3D model properties, irrespective of any other external quantities. The algorithms created in the OTP database that are based on the product technological code, with the inclusion of specific data from the retrieved technological dataset, also result in a precise binary classification for manufacturing. Data entry and all database algorithms are carried out in a fully automated way. The research has fully validated the set hypothesis. Taking into account the previous studies [1]; [28], it can be stated that a new scientific method has been devised for designing basic technological processes with the aim of rapid, 3DM-based product cost assessment. The method is called the OTP method, and its further development will pursue the level of detail in coding the manufacturing process. The achieved binary classification can be further elaborated by including more details about the elements of the basic technological process. Decimal and special technological classifications would be developed.

Item Type: Thesis (Doctoral thesis)
Uncontrolled Keywords: projektiranje tehnoloških procesa; CAD 3D model; tehnološka klasifikacija; kodiranje izradaka; procjena cijene koštanja; osnovni tehnološki proces (OTP); aplikacija za dohvat značajki (BTP); zapis tehnoloških značajki (.btp); OTP baza podataka (OTPbp); OTP metoda (mOTP)
Keywords (Croatian): process planning; CAD 3D model; export features from CAD 3D models; technological classification; encoding of parts; production coding; cost estimate; basic technological process; feature retrieval application; record of technological features; OTP database; OTP method
Subjects: TECHNICAL SCIENCE > Mechanical Engineering
Divisions: 700 Department of Industrial Engineering > 710 Chair of Production Design
Date Deposited: 25 Jan 2018 08:51
Last Modified: 25 Jan 2018 08:51
URI: http://repozitorij.fsb.hr/id/eprint/8232

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

Nema podataka za dohvacanje citata