Community Research and Development Information Service - CORDIS

Final Report Summary - CLAMPIT (INTELLIGENT WELDING CLAMP DESIGN SOFTWARE USING COMPUTER-AIDED OPTIMIZATION FOR SMES TO ACHIEVE HIGH PRECISION ASSEMBLY)

Executive Summary:
A fundamental part of all types of manufacturing technologies is clamping. This operation is critical especially in welding where the high temperatures and forces due to possible generating deformation therefore additional costs. These costs such as fabrication cost to reduce the distortion, material costs and of course the time spent by correcting or reproducing the workpiece are generating high pressure on especially smaller welding workshops in and outside the European Union.

In the frame of the CLAMPIT project (Grant agreement no.: 315208) the members of the Consortium target the development of a high-end decision support system that helps metalworking SMEs carrying out welding work to increase their efficiency by aiding the preparation of fixture design. This CLAMPIT application provides assistance to find the optimum fixture design containing optimal clamping points and forces. In addition it includes a fast and efficient distortion prediction module incorporating LSND technology.

The concept behind the CLAMPIT software is proven basically in every day’s life. If a new problem appears for a person he tries to solve it by himself or asks for help from others that have experience facing similar problems. The situation is the same in designing a fixture plan for a new problem: the welding engineer prepares the clamping design on his own or asks from help from an experienced welding expert. In CLAMPIT both the previously mentioned scenario can be achieved: the software gives an interface to the welding engineer where fixture plan can be prepared by sketch or the welding designer can ask for assistance from the software. After defining the given problem, the application suggests a list of potential solution to the user who selects one from the list and can adapt it to the given problem. The fixture plan preparation is supported by a distortion measurement incorporating LSND technology which can simulate the welding effects on the workpiece and can measure potential distortion before the actual welding process sparing lots of time, cost and increasing the quality of the welding process. The newly customized fixture design can be saved back to the knowledge base, so the system is a “learning-by-using” system: the more the application is used the cleverer the system gets.

Project Context and Objectives:
The manufacturing and engineering industry depends heavily on workflow efficiency. Welding is an essential area of industrial manufacturing and smaller welding workshops are under economic pressure to find new solutions in order to stay in a competitive market, while the number of skilled welders decreases every year. Clamping plays major part of numerous types of manufacturing, but especially in welding operation unwanted distortion can be appear due to forces and high temperatures. The task of specifying clamping positions and forces for welding assemblies are among the most common everyday tasks for welding SMEs. At present preparing the optimal clamping design for a simple design needs couple of hours of an engineer or an experienced technician, a more complex assembly containing 3-5. Based on the internal investigations of welding expert SMEs of the consortium, a small size company is estimated to allocate 50-600 engineer hours a year to prepare clamping designs.

The CLAMPIT project aims at increasing the competitiveness of a large group of SMEs by developing an innovative decision support system to reduce time and cost needed to prepare fixture design and increase its quality. The application assists its users to find the optimum fixture design containing optimal clamping points and forces. In addition it includes a fast and efficient distortion prediction module incorporating LSND technology.

One of the main advantages of CLAMPIT is the usage of the previously collected and stored experience during designing a welding fixture plan. For this purpose case-based reasoning (CBR) approach is used that helps the welding designers by referencing previous design cases, to make fixturing solution quickly. The newly created cases can be saved back in the case base extending the knowledge of the system. The generated cases can be later converted into a structured format by the assistance of the report generation module. CLAMPIT will include an additional module - Low Stress No Distortion (LSND) module - that can be used independently, together with clamping optimization (CBR), or can be excluded depending on the needs of the user. This module aims to apply thermal methods (LSND techniques) in order to reduce the residual stress and distortion in welded parts. The developed module will be able to optimize LSND conditions for a component based on the CAD model and the Welding Procedure Specifications. The introduction of CLAMPIT will speed up the methodology of the welding process as the whole clamping design can be done quickly and accurately saving time. The use of case based reasoning (CBR) to design clamping fixtures has been well documented and this can be applied with other decision support systems (DSS) to give accurate information (visually as well) to the welding engineer making the work more efficient. It is more and more accepted that knowledge, experience and expertise can be represented digitally by the use of computer systems in special cases. CLAMPIT will use CAD file recognition and CBR to supply the knowledge and experience which would take years of training. It is important to note that the system will not replace but enhance the job of a welding expert giving maximum benefit and reduce the possibility of error. With the developments in the information technology sector the CLAMPIT consortium had realized that this existing need could had been solved by targeting the following objectives:

Scientific objectives:

• The principal scientific objective is to explore the forces acting on a weld with respect to geometry and heat input with the help of an in depth literature review and in-house experiments in order to investigate the following parameters:

o Amount of heat
o Heat conduction area
o Parent material (heat conductivity)
o Location (with respect to geometry)
o Residual stresses generated by the heat

• To explore how the stresses (due to the heat gradient) affect the geometry distortion (permanently) and to investigate minimal clamping forces to eliminate distortion (to overcome stresses). The information gained will be studied together on the effects of using Low Stress No Distortion (LSND) method on weldments.

Technological objectives:

• To design the structure of CLAMPIT software and database based on the system specification which is defined as the outcome of WP1 based on the market survey results. The database and software design will result in a schematic plan of all the necessary modules, their functions, inputs, outputs and their interaction.

• To develop a database that is capable of storing information about the geometry, the welding processes and clamping equipment models available on the present market ensuring that the database can be updated in a simple way if new types appear on the market

• To develop a decision supporting system that is based on neural networks and contains all the welding process relations. The software will calculate stresses and required forces based on: Welding Procedure Specifications (WPS); Geometry; Assembly; and Stress factors – eg: LSND. The information will be used to determine the optimal clamping position by taking into account the company’s available clamps and tables stored in the database.

• To develop a CAD model handling module in which specifications of a welded product can be graphically inserted. The CAD module will be able to manipulate complex assemblies as well as the database of existing welding fixtures.

• To investigate the effects of low stress no distortion (LSND) in order to reduce plastic strains in the weld zone, providing a weld with low residual stresses and minimum distortion. The LSND development in WP6 led by TWI will allow the software to combine the fixture design with stress reduction ensuring a better quality weld. The modules designed will be suitable for integration into commercially available clamping systems

• To develop a graphical user interface (GUI) for CLAMPIT system that is suitable for industrial use. The GUI will also be the platform where the different modules are integrated and a beta version of CLAMPIT with the following functionalities will be presented: CAD recognition; Force calculation; Clamping fixture specification; and LSND information.

• To perform tests with CLAMPIT system and validate its functionality in industrial conditions at the end users’ (ROBOTEC & BARIDA) facilities. The validation will involve all partners and will be done using past welding project information to compare the increase in quality and time saved. Further tests will be carried out on actual welding tasks.

• To certify the CLAMPIT technology and the resulting new fixture determination software by a recognised body ensuring that the quality and the standard reached is industry applicable.

Market and business objectives:

• To develop an on-line service based on CLAMPIT that is available to even small and medium sized companies such as welding table manufacturers or suppliers and welding workshops.

• To successfully train the SME participants so they are fully in the position to use the developed technology. They will be trained through regular meetings and training events which will be held before the validation.

• To prepare a “Tutorial” and a “Best Practice Guide” included in the user guide of CLAMPIT.

• To prepare the ground for a successful exploitation of the developed prototype in the post-project phase through extensive dissemination and networking activities. This will be achieved through conferences, publications, seminars and a multi media guide amongst other activities.

• To prepare a viable and implementable business plan for participating SMEs based on which they will be able to benefit from the project results while protecting intellectual property rights arising from this project according to their needs.

Project Results:
The project was two years in duration and it started on the 1st of September 2012. The main result of the whole project is the CLAMPIT system which is a decision supporting tool for welding experts. It provides assistance in designing fixture plans for welding processes resulting significant time and cost saving for the welding SMEs. The CLAMPIT system includes (I) Decision Support System based on Case-based reasoning (CBR) (II) Distortion measurement incorporating Low Stress No Distortion (LSND) technique (III) Knowledge base stored in the developed Database. The present description introduces the public information on the steps how the final output of the project was realized.

The work started with preparation of the system specification based on the relevant market needs. The market was examined from two sites. On one hand an online questionnaire was prepared in 6 different languages (English, Hungarian, Slovak, German, Spanish and Turkish). It contained 36 questions organised into four groups: Company profile, IT related questions, operation and technology related questions and company details. The answers were collected continuously and the results were analysed. On the other hand, the market needs were also analysed, based on the experience of consortium partners and on available public information. Product research was carried out to identify as many solutions and products that facilitate design of welding fixtures as possible. Although there were found several methodologies and research works elaborated on the topic of computer-aided fixture design (CAFD), no available software applications were found on the market that would be dedicated to the welding fixture design in the same way as CLAMPIT is capable. Literature research was done as an overview of the publications related to the project where the collected information helped in the development of CLAMPIT. The literature review was carried out on the following topics: (I) Computer aided fixture design, (II) Shape retrieval using 3D descriptors, (III) Low Stress No Distortion (LSND). Based on the information received from potential end users, literature review, market information about commercially available systems, from the consortium partners’ expertise and their own requirements the functional system specification was outlined. It contained the conclusions of the requirement analysis, constraints of the system, functional requirements (module requirements, user handling), non-functional requirements (performance, data definition, user documentation).

The work continued with designing the software architecture and database structure. The software design contained 5 layers: (I) Presentation layer as user interaction layer, (II) Knowledge layer consisting of three distinctive parts. Case Based Reasoning (CBR), Low Stress No Distortion (LSND), Shape descriptor modules, (III) Data structure layer that plays part in data formatting and sorting between the software components and the database, (IV) Data layer which is responsible for the physical storage of the data, (V) Administration layer to be in charge for the maintenance mainly of the database structure. Database design contained data description how the major data or system entities are stored, processed and organized. The database tables can be logically grouped into three different categories: (I) Fixture plan related tables, (II) Shape storage and retrieval related tables, and (III) additional tables such as users data table. As a result of these activities the first milestone of the CLAMPIT project about developed software and database structures was successfully achieved.
Once the design was finalized, the database development started. A database server was installed and configured at which was followed by the implementation of the database tool. This tool was used by experts and developers to upload necessary data and parameters to the database without having deeper knowledge of the DB. The necessary data was provided by creating the assemblies in real life, testing the fixture plans. With the agreement of the consortium the data uploading process was continuously performed until the end of the project, so tests contained lower complexity workpieces in the beginning, and more complex examples were prepared and uploaded later. Number of weld types were limited (only stainless steel material is built into the module) and only linear weldments were used. As a result more than 70 workpieces with ~100 setups and ~110 fixture plans were uploaded in the knowledge base.

As next step, the development of the Decision Support System (DSS) started. The DSS is using Case Based Reasoning (CBR) approach that helps the user to find a possible solution based on the solutions of similar past clamping problems. This approach is based upon the assumption that similar workpieces require similar fixture designs. Semantic representation of the mechanical meanings of a fixture plan and different clamping equipment was defined. As result abstract objects and relationships among these objects were generated. After the knowledge interpretation was successfully carried out, the implementation stage started by declaring the previously defined objects and relationships into .NET based classes in C# programming language. Next step was to implement the core algorithms of the CBR such as the similarity function which plays majestic part of the case retrieval process. Two types of testing framework were developed to test the developed algorithms. The first version contained only binary data, while the second testing framework contained also the support for 3D CAD files (.STEP file format).

In order to compile the used CAD file format (.STEP), visualize results in 3D and to create specific algorithms that are able to find characteristic features of a model, CAD module was developed. The development contained customized algorithms for the 3rd party software library, DevDept Eyeshot and implementation of the 3D shape descriptor that plays major part in the comparison of the shapes. As next step the output of the similarity function calculation was integrated into the DSS module.

Researchers analysed two type software models (thermo-elastic-plastic and shrinkage method) and as result the distortion prediction module based on the shrinkage method was designed and developed. It is able to perform on-the-fly finite element calculation by assigning anisotropic thermal expansion coefficients to the model.

To hide complicated processes from the user and to display intelligible results, simulated environment and Graphical User Interface (GUI) was developed incorporating all the previously developed modules into one interface.

All the separate parts were integrated together and passed several initial laboratory system tests before sending the software out to the partners. The acceptance tests were carried out by the SME partners supported by the developers. Based on the test results the system had positive effect. During the period the state of the art was continuously monitored, and information was shared within the Consortium.

As a summary, the main results achieved are as follows:

• The CLAMPIT system was fully designed with the assistance of SME partners reflecting the content of the system specification

• Best shape comparison method was selected

• All parts of the CLAMPIT system constructed and successfully integrated

• The system was tested and validated on laboratory scale and at the end-users as well

• The capacity of the CLAMPIT system reached the targeted level defined in the system specification

• Multiple distortion measurement approaches were analysed and the best was applied

Potential Impact:
Welding technologies and welding consumables are representing a potentially high growth lately, especially when considering the significant revenue being generated by the segment of the global welding consumables market. With the fast developing technological knowledge on robots, the welding machinery industry can also highly benefit from the world of more sophisticated automations. The appearance of robots, laser-systems and computer-aided manufacturing systems is surely re-defining the prospects of all industries across the world.

Welding is regarded as another metal processing method in the 2009 edition of EuroStat Statistical Book „European Business”. Along with other metal processing activities, this sector was the largest in terms of enterprise numbers, persons employed and value added. Activities grouped within this sector were the principal activities of an estimated 158 000 enterprises, providing employment for 1.4 million people, about one in every four (28,2%) of the metals and metal products workforce.

Welding plays a significant role in the automotive and transport industry. Transportation and transport equipment manufacturing sector is vital to the economic development as it provides the nodes of transporting both individuals and more importantly, moving of goods that establish the basis of trading and market. Demand for transport equipment has risen as the volume of goods transported and the distance travelled by passengers have expanded immensely. The EU’s transport equipment manufacturing sector consisted of 45.7 thousand enterprises which employed 3.2 million persons in 2006. Paid employees dominated this workforce, accounting for 98.6 % of all persons employed: this proportion was above the non-financial business economy average (86.5 %) as well as the industrial average (94.2 %) and in most of the subsectors the proportion reached 99.0 % or higher.

Clamping is an initial step in most of welding processes no matter if we talk about manual or automatic welding. The optimal clamping method ensures that the individual work pieces will be joined in the correct position and the optimal forces and accessories contribute to avoid distortions that can occur during welding due to the impact of heat. From this perspective the design of fixtures is an essential operation during the welding process and therefore requires highly experienced welders for every new assembly. Mistakes in the design can delay the actual manufacturing operations and thus reduce productivity of welding workshops.

CLAMPIT is an assistive software which can be perfectly used in the designing work segment of the welding industry. Currently, in the prototype version, few boundaries were introduced such as: the LSND distortion measurement can be applied only for a few types of welding methods, and only straight welds are supported as well as only stainless steel material is built into the LSND module. However, thanks to the “universal” design, these features can be further improved by extending the system with new welding techniques and a kit of tools.

The optimal clamping designs for welding should avoid the errors in the structure building and also the distortion caused by the heat while applying the weld. Nowadays preparing the optimal clamping design requires significant human resources both in time and experience. As explained in the DOW a small size company is estimated to allocate 50-600 engineer hours a year to prepare clamping designs. It is estimated by the consortium partners that the CLAMPIT software would be able to save from 10 000 to 25 000€ per year in a medium or small metal company regarding the save on human labour costs and the estimated losses of the company derived from avoidable distortion exceeds.

The key features of the developed CLAMPIT software – compared with the currently available main competitions on the market - are:

• Re using the “knowledge” of the previous cases:

Case-based reasoning is used in CLAMPIT to help the design process of a fixture plan creation. When a new clamping problem appears, the user first loads in the CLAMPIT software a 3D CAD file of the given work piece to be welded. The system interprets the model and identifies important parts. After the user extended the feature/functional requirements the Case Based Reasoning (CBR) offers possible solutions to the user by finding similar cases from the previous clamping problems stored in the database. The CBR search can be re-executed any time later with fine-tune parameters in order to find the best-fitting possible solution. The Case-based reasoning approach is based upon the assumption that similar workpieces require similar fixture designs. The module’s performance depends on the database size of course; as the number of the stored cases constantly increasing, the higher the chance that the module finds a similar case and therefore can suggest a valid solution for a new problem.

• Fast distortion measurement calculation:

As measurement simulations usually require lots of processing time, a new aim was set towards the LSND module during the specification: provide relatively good result within significantly less time than current software packages. In many cases the user needs quick results; regardless the outcome is less accurate. The more time needed for the simulation the more accurate the result measurement is. In addition the developers introduced a scale with more approximation levels (in each level the original CAD file is approximated with different amount of mesh elements), so the user can select according to its needs/computer capacity the approximation level in the measurement of distortion.

• Creating / building a knowledge base of welding processes:

Storing each case in the database creates an ever expanding knowledge base of welding fixture designs. This “knowledge” can be easily transferred between CLAMPIT databases as data packages, therefore new market possibilities open by trading this part of the system. For example a company creates a database of 500 cases of specific type of models, which can be bought as an extension for CLAMPIT for additional charge.

The CLAMPIT software offers to the market a new capacity for developing welding structures and solutions. The benefits for the metal welding market are:

• Improve the clamping design: it means producing high standard quality products with a better applied system of welding and also avoiding the distortion of the final product. Clamping the structure is the initial phase in the welding process. Assemblies should be placed in a clamp for manual or automatic welding. The optimal clamping method ensures that the individual work pieces will be joined on the correct points and arcs. Of course an optimal clamping design using optimal forces and accessories contributes to avoid the distortion during welding.

• The correct clamping design ensures that the final welded product would reach high quality standards.

• Reduce the waste of time and material produced by the errors occurred with the traditional clamping system. It is common in the metal welding industry to have structural dysfunctions. Structures which do not reach the quality standards should be rejected. This rejection supposes a huge waste of time and material that would be avoid by applying the new software.

• Reduce the human/hours allocated on preparing the clamping design. Nowadays, clamping process requires significant resources both in time and experience. While a simple design requires a couple of hours for an experienced engineer, a complex assembly might require much time for calculation. An engineer should calculate the safest position for work pieces that allow using the appropriate welding technic and studying the necessary clamping accessories to be used. A small size welding company is estimated to allocate 50-600 hours per year to prepare the clamping designs. A case study is described in the DOW that proves with optimized clamping software the expected savings per year would reach 25.000€.

CLAMPIT will be exploited by the SME members of the consortium. After the project end there will be three phases:

1) PHASE A: Notwithstanding the current software has to be upgraded the partners consider that they actually have a real product that can be used in their own business and sold to third companies.

2) PHASE B: Product/software upgrading. The SME partners aim to further develop the product in order to reach a ready to market product that could fill the end-user expectations.

3) PHASE C: Once the partners end the upgrading phase the result will be a market ready software that would be launched by a business operating joint ownership.

The SME members are allowed to use the software as end users for their own business. In that case the software is used on a royalty-free basis. Preliminary, BARIDA and ROBOTEC will use the software for their own purposes. The software will be distributed via contracted third partners (external distributors). This strategy will allow the rapid and easy introduction of the product into the market. All the partners during the PHASE A are allowed to find distributors and sell through them the current product. A new commercial web page for the product is the main communication channel toward the market. There will be information about the owners SMEs and their distribution territories. Any of the owner SMEs will be able to distribute the current product within their territory independently.

A number of dissemination activities have been carried out in the second period which is detailed in D9.8. The two project videos and the project website remained the main dissemination tools of the project. Besides the project website, NEMETSCHEK created and designed the product website as they have vast experience in establishing and operating websites. For the request and initial plans of the consortium, the official web page of CLAMPIT contains all the product information; information about the project and the Consortium; specialized information about the sector; section for on-line purchasing. Newsletters were issued with regular updates on the project status which were uploaded to the project website. The project leaflet was circulated at several events in different languages. ROBOTEC and INTERN attended the World Welding Fair in Essen, Germany; while PZVAR regularly attended the other reputed conference, the International Engineering Fair in Brno, Czech Republic. ATEKNEA participated in the 9th International Joint Conference on Software Technologies in Vienna, Austria with project presentation and further represented the project at ELTE Innovation Day, INNOSKART ICT cluster meeting, Innovation Conference organized by ATEKNEA in Budapest. BARIDA represented CLAMPIT at the 6th Automobile and Commercial Vehicles, Component and Spare Parts Fair. Each consortium member contributed to regularly updating the target public audience on the actual evolvement of the project via website news and short articles, such as the article on CORDIS. For further dissemination, the demonstration software was created by the partners.

List of Websites:
Public website address: clampit.eu
Contact details: clampit@ateknea.com

Related information

Reported by

ATEKNEA SOLUTIONS HUNGARY KFT
Hungary
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top