Community Research and Development Information Service - CORDIS

H2020

RADICLE Report Summary

Project ID: 636932

Periodic Reporting for period 1 - RADICLE (Real-time dynamic control system for laser welding)

Reporting period: 2015-02-01 to 2016-01-31

Summary of the context and overall objectives of the project

1 Summary

The overall aim of the Radicle projects is to develop a laser welding adaptive control system which can integrate sensor data from 3 loops in real time by adjusting laser parameters to deliver welded joints with zero defects. The control system will include in process monitoring control, fault prevention / fixing pre and post welding measurement.
Radicle project partners include four large, end user partners which includes Roll Royce, GE, GKN and CRF (Fiat). The successful implementation of Radicle technology through Project Partners and the wider industry, will enable the project to have the following impacts:-
• An increase in Health & Safety benefits
• An increased productivity of up to 30% resulting in reduced emissions, reduced energy usage, reduced rework and a reduction in the need for final NDE testing
• Reduce floor space requirements by removing the need for large welding room enclosures

The Radicle project contributes to the wider European 2020 targets by:-
• Reducing energy usage and greenhouse gas emissions
• Increasing employment for 20-64 year olds
• Increasing R&D spend
• Increasing education, particularly third level education

Laser welding is a high performance joining process which provides significant benefits over more conventional methods, such as arc based welding. The global market for using laser equipment for material processes has grown 50% since 2004 and is valued in excess of €10 billion per annum. The automotive industry has led in the implementation of laser welding for structures such as BIW (Body in White). The aerospace industry are now also using this process for various applications in joining reactive metals such as titanium alloys. Welds in the aerospace undergo 100% post weld inspections using X-ray technologies and welds with greater levels of porosity than relevant standards specify, have to undergo rework which can severely impact the economic benefits of laser welding processes.
There are 3 current levels of inspection and control used before, during and after laser welding to assist in reducing weld defects. These include:-
• Loop 1 – Seam tracking and pre-process adaptive control
• Loop 2 – In process monitoring
• Loop 3 – Post processing welding NDE
Loop 1 will be linked to the adaptive control system which will enable self-optimisation of the pre-process control algorithms during welding. The data generated from demonstration and validation activities will be used to develop initial pre-process control algorithms, specifying the actions required to be taken when certain joint geometries are met.
Loop 2 will develop novel plume analysis sensors. The laser generated plume contains a number of variables which relate to the immediate conditions within the keyhole region. The mechanism of formation of these by-products and the material phases reached during processing alter as a consequence of process conditions.
Radicle will analyse the fume during collection and with techniques such as forward optical scattering, these fume process parameters can provide additional inputs for the adaptive control system. The Radicle project will also investigate current sensors which use laser light to measure weld pool depth to determine if further information can be obtained from the reflected laser light.
Loop 3, the Post-process NDT/NDE characterisation of components, will be used to inform and update the parameters database within the Radicle system. This will match process conditions at the time of the weld, with data collected from sensors on Loop 2 and the Loop 3 NDT/NDE results. This will assist in identifying patterns which contribute to defects, enabling the algorithm to “learn” common processing issues or defect forming parameters.

The RADICLE system could also potentially be linked into the surrounding IT infrastructure and Manufacturing Execution Systems [MES]. This would offer major benefits in the aerospace industry as a component could be traced back to the parameters used to weld it. While this may be beyond the scope of RADICLE it will be considered throughout the project to ensure the algorithm is compatible with any future developments in this area.
The RADICLE project combines expertise in welding, materials, optical measurement, lasers and software/ICT. It is also crosses industries in that, solutions to the power, automotive and aerospace sectors can be offered. The project brings academic and industrial partners from across Europe together, benefitting the partners and the project as a whole.

Figure 1: Overview of the Radicle Concept

The MTC is consortium manager of the Radicle project and technical work has been managed using 6-monthly work plans, second level plans and 4 box reports.. WP7 is led by EWF who manage the dissemination of information to the broader welding community and also maintain the business plan on behalf of project partners to ensure the project results will be rapidly commercialised.
Radicle is a 36 month project and is delivered in three distinct phases. Phase 1 is about the specification and preparation of samples and components for initial characterisation and subsequent testing. Phase 2 is the development of sensors and the algorithm whilst phase 3 is the integration, demonstration and validation of the RADICLE system, firstly in the lab and subsequently on MTS’s laser welding cell.

Figure 2: Overall approach and methodology

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

3 Explanation of the work carried out by the beneficiaries and Overview of the progress

Work carried out during the first twelve months of the project included objectives from Work package 1, 2, 3, 8 and 9.

Work package 1 (Starts Month 1 and ends Month 3): is 100% complete and all deliverables have been submitted. Under Work Package 1, the following activities have been completed:
Five end-user applications were selected:-
• Structural cases and housings for gas-turbine aero-engines, vane assembly for power generation gas-turbines, automotive body structure assembly and general engineering structural applications
• Each end user application has been described in detail to provide a justification of why it was selected for evaluation within the Radicle project. The description includes justification of the Radicle control system with respect to the specific end-user application:
• A detailed breakdown of the application itself including materials, weld configurations etc.
o Production volumes.
o Process considerations (shielding requirements, potential defects, material preparation)
o Possible rework strategies.
o Tooling considerations.
o Process parameter capability and robustness (the ‘process window’).
o See Deliverable D1.1 for more detail.
• Specifications for the required weld quality for each end user application with reference to:
o Recognised standards from international bodies.
o Standards internal to the end-user.
o Where applications were prospective (gas-turbine vane assembly, for example), the consortium (particularly TWI) was able to suggest weld standards appropriate to that application.
o See Deliverable D1.2 for more detail
• A discussion for the requirement s and content of the specifications for the welding system for appropriate for each end-user application.
o This does not necessarily mean that a complete specification as presented for each application, but a list of specification considerations was presented for each application.
o See Deliverable D1.3 for more detail.
• A discussion and description of the sensing and monitoring requirements for each application including:
o Definition of the process monitoring / control ‘loops’ discussed in the Grant Agreement document.
o Potential process outputs that could be monitored.
o Work package 4 development.
o Initial sensor specification.
o See Deliverable D1.4 for more detail.
The outputs of this work package were supported by an extensive review of appropriate literature, corroborated by experiential learning from the consortium partners wherever possible.

Work Package 2 (starts month 4 and ends month 9): focused on the definition of test sample specification, production and testing activities. An approach for producing and characterising laser welded samples was developed and a testing methodology was created, comprising a series of non-destructive and destructive inspections tests, to analyse quality and properties of the produced welds. Produced reference welded samples and also with known imperfections (e.g. porosity, incomplete penetration, etc.) were used to evaluate commercially proven in-process monitoring sensors, to benchmark their capability for Loop 2 activities. Furthermore, an initial approach for the detection of key process variable and tolerances within which these variables must ultimately be contained by the RADICLE control system were identified.

Radicle partners actively worked together to effectively cover the activities described above. WP2 started at Month 4 of the Radicle project and was scheduled to be completed at Month 9. However, Task 2.3 was completed at M14 instead of M7 of the project, due to the need to include validation of two additional sensors setups (e.g. external photodiodes and laser vibrometer), compared to the original list. Importantly, it should be addressed that this extended period is not expected to have an impact on activities in WP5 (starting at M18) and/or any other activities of the other WPs. Furthermore, regarding Task 2.4, additional activity is proposed to map and optimise processing, to support the development of the initial process control sub-algorithms for each of the case study parts. This work constitutes an additional activity to the current work programme and, if approved, it is expected to be completed by M16 of the project.
Work package 3 (starts month 4 and ends month 18): is progressing through the collaborative efforts of the technical delivery partners (and the wider consortium community) and has so far generated outputs that are not only of value to the Radicle project, but also to the wider engineering sector.
Many potential in–process sensors have been tested during the practical work being undertaken to support WP2, which forms the foundation work for the knowledge base required for Task 3.3 and the deliverable D3.5:
• Photodiodes:
o Close-coupled and remote detectors.
o Used by a number of commercially available process monitoring systems.
• Spectroscopy:
o Commonly cited in academic literature.
• Particle detection.
o Previously tested for alternative processes.
o Cited in academic literature.
• Acoustic monitoring:
o Previously tested at high TRL level by Rolls-Royce for an alternative process.
o Commonly cited in academic literature.
• Non-contact vibrometry:
o Considered academically to overcome some of the issues inherent to, and limiting acoustic monitoring.
• Bespoke plume analysis systems (task 3.2):
o Still under development (as per task 3.2).
o A bespoke test rig has been built to maximise concept opportunity.
o Sensors suites appropriate to the application have been identified and incorporated.
o Currently only capable of post-event analysis therefore real-time correlations are currently unknown.
As a result of the above testing work, two sensor signal gathering systems adopted – one commercially available, the other a bespoke unit developed by LOE for the Radicle project. The sensor data gathered from the above work, combined with measured weld attributes will provide the input necessary for the WP4 algorithms. Initial data has already been passed to VTT for analysis and review.
Camera systems have not yet been evaluated but will be tested during further work required for WP2 (see section 5: Deviations to Annex) undertaken by TWI and MTC. LOE will support these activities. If successful, camera inputs will be included in subsequent iterations of the signal collection system described above.
A number of key challenges have been identified as a result of the work undertaken for WP2 and WP3:
• There is no synchronisation between multiple sensors yet.
• No synchronisation between sensors and process yet
• Correlation between sensor signals and the resulting weld attributes not yet established. However, solutions are under development for incorporation into future work for WP 2 and 4.
The deliverables completed to date include:
• D3.1
o Close collaboration with consortium partners to identify measurable weld emissions and attributes.
o Collaboration with TWI and MTC in undertaking practical work to establish sensor performance.
o Evolution of the Defects Vs Sensors spreadsheet.
• D3.2:
o Includes the specifications for the sensors required for the plume analysis sensor being developed under task 3.2 and required for deliverable D3.4.
Deliverable D3.3 is still work-in-progress. An initial draft has been completed, building on the work undertaken to support deliverables D3.1 and D3.2.

Work Package 4 (starts month 4 and ends month 18): has been focused on tasks 4.1 and 4.2 during the reporting period. Task 4.1 “Specification of hardware and software systems” has specified an integrated machine intelligence architecture for the adaptive control suitable for identified weld control challenges. The total completion is 70 %. Task 4.2 “Development and integration of control algorithms” has focused especially on stacked denoising auto-encoders, tile coding algorithms and Q-learning. The total completion is 50 %.

Work package 7 (starts month 1 and ends month 36): are the dissemination, exploitation and communication activities for the project and during the first 12 months of the project, is 20% complete. As this work package is run over 3 years, we expect most of work to be carried out during the remaining 24 months of the project. Project partners have demonstrated great commitment to these tasks having been involved in over 40 activities. These included participation in conferences, face-to-face meetings, magazine articles, flyers, posters and other type of activities. The Radicle website was set up according to schedule and currently presents an average number of visitors per month of over 100. Also, as it is clear in Deliverable 7.5 all partners have been involved in dissemination activities and have reported them to EWF (as WP leader) every 6 months.

Work package 8 (starts month 1 and ends month 36): is the Consortium Management of the Radicle project and has been led by the MTC. The MTC have ensured regular communications in the form of monthly teleconferences, work package kick off meetings, 3 monthly technical meetings and 6 monthly consortium meetings, take place. The MTC has supported partners in approval processes, interim reporting, facilitated contractual changes and owned consortium’s NDA’s with third parties. The MTC is facilitating the full integration of our new partner, GE Power into the Consortium Agreement Partners have worked well together with and data has been shared i.e., Rolls Royce has presented the results of titanium welding gathered from a previous project.

Work Package 9 (starts month 1 and ends month 36): is the Scientific and Technical Management of the project. Led by the project’s technical lead, the project has been monitored using detailed level 2 plans derived from the Radicle task and deliverable lists. These plans are used as a working document and are reviewed and updated in an ongoing, monthly process.
The monthly project monitoring process involves:
• A monthly teleconference, hosted by the MTC, for all consortium partners.
• A 4 Box report, updated monthly, required by any consortium member undertaking work package activity within that month and presented / discussed at the monthly teleconference
• Review of the risk register (see below) if required.
A technical meeting has been introduced to take place once every three months, requiring the representation of all consortium partners, preferably face-to-face. Every second meeting will correspond with the day before the required 6-monthly Consortium meetings. These meetings are for the discussion and resolution of technical issues and questions, including exploitation plans.
Monthly, informal meetings have also taken place involving the UK-based consortium members (MTC, TWI. LOE and Rolls-Royce) with the aim of identifying, discussing and resolving any issues regarding the delivery of the practical data-gathering necessary for the development of the Radicle system. Other non-UK consortium partners and key technology suppliers have joined these meetings either in person or via teleconference as and when necessary. Any partner can be involved in these meetings on an ad hoc basis as and when required.
A new kick-off meeting is also being introduced for each upcoming work package, and this will be hosted by the work package leader and should involve all consortium partners, particularly those involved with the activities from that work package. One of the outputs from this meeting is an outline second level plan for the delivery of the task and deliverables required within that work package. This outline can then be deployed into the master second-level plan.
A Risk Management register has been created using an industrially accepted format and is now live on the Radicle website. It is available for editing by any consortium partner.
Further details of work package progress, Consortium

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Environmental and Societal Impacts

With 12 months of work the RADICLE consortium has been able to identify several Societal and Environmental Impacts adding or complementing the ones described in the proposal. These impacts will have a great impact in the manufacturing sector and on Europe as a whole, are outlined below:-

• RADICLE will help to increase economic growth and create new jobs in Europe by means of reducing costs in manufacturing for major employers in Europe. Increasing levels of business by being more price-competitive, will result in higher employment rates. In the short term, the main benefit will be with the industrial partners in the consortium but will later be expanded to other manufacturers, including SMEs in EU states, via the EWF. Adding to this and based on the initial development of the RADCILE business plan, it is clear that the benefits and economic growth will not just be limited to end-users (both SMEs and Large companies). One of the possible exploitation routes would be in collaboration with companies that already are in the market selling laser equipment and sensors. The RADICLE system will allow them to gain a competitive advantage when compared to the competition and increase their market share while at the same time, implement laser welding systems across Europe which would provide the end-users an advantage in terms of manufacturing capability and quality.
• Increased output of the OEM manufacturers will have a positive effect for their supply chain, targeting not only large companies but also SMEs. By having large OEM manufacturers in the consortium we can ensure that the RADICLE system is transferred through the entire supply chain of these OEMs.
• Improved Health and Safety as the operator is remote from the laser welding head and therefore, not exposed to fumes and potential injury. Adding to this and related to the fact that the RADICLE system will allow the end-users to obtain better quality welds, it is expected that the number of samples/products that need to be re-worked will decrease considerably. This will also result in improved health and safety. Furthermore, the incorporation of LOOP 3 in the RADICLE system will also reduce the exposure of workers to, in some cases, high/medium risk nondestructive testing (for example x-rays).
• Training and education opportunities for the partners to upskill their workforce and train other people working in the sector. The training and education gained more relevance in the RADICLE project, since it was seen as a way to increase the visibility of the RADICLE system as well as a way of ensuring a proper uptake of the system by industry. This approach moves from an “in-house” training to a sustainable training/educational program that will be included in the EWF training system that is already running for 25 years.
• Increased employment within the laser systems supply chain to design, manufacture and install the new equipment and sensors. This increase of employment is not only related to what was mentioned in terms of an increase of sales of equipment (more information in D7.4) but also to the increase of well trained workers that will be achieved through the development of training guidelines that will, at a later stage, be incorporated within the EWF training and qualification system.
• Development of gender-agnostic employment opportunities as laser-welding does not involve heavy manual labour effect on the EU population due to the reduction in pollution. By reducing the need for machining and by enabling the increased uptake of laser welding, the working environment will be improved as, for example, workers will be less exposed to the hazardous metal working fluids (MWFs) used in this type of process.

Energy and Environment: The main benefit to the EU outside of the direct economic benefits will be in the reduction of energy consumption. In 2004-05, 28% of the EU-27’s energy consumption was by industry; and 20-40% of the energy used in industry is of no added value. As shown above, the RADICLE technology has the potential to reduce energy usage by ~125MJ/kg, equating to potential CO2 emissions saving of 15,600kg (CO2) / kg (part). Assuming a modest output of 1,000 units per year (20kg parts) this would equate to total energy savings of ~2,500,000MJ/year with CO2 savings of 312 tonnes/year. In addition, there is potential to reduce weight for automotive and aerospace parts which again will lead to great energy and emissions savings during lifetime.

Related information

Record Number: 190221 / Last updated on: 2016-11-09
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top