Skip to main content

Development of Short Distance WELding Concepts for AIRframes (WEL-AIR project)

Final Report Summary - WEL-AIR (Development of Short Distance WELding Concepts for AIRframes)

The WEL-AIR consortium systematically investigated the basic mechanisms of hot tearing control, crack initiation and crack growth at the vicinity of the run-in and run-out of skin-stringer welds as well as short distance clip-skin welds. Therefore, the WEL-AIR project provided new knowledge for:
- innovative fabrication and design concepts for clip-skin configurations;
- design or welding recommendations to avoid hot tearing;
- development of new joint configurations using Laser Beam and Friction Stir Welding processes;
- development of improved understanding of the damage tolerance behaviours (fatigue, residual strength and corrosion) of short distance welds.

For strengthening the competitiveness of the European aeronautical manufacturing industry, knowledge based development of high performance and innovative metallic light-weight airframe concepts are important. For this purpose, weight and cost efficiency will be reached with the development of the 'Integral Structure' or 'Rivet-Free' Al-alloy airframes through the use of advanced welding technologies such as laser beam welding (LBW) and friction stir welding (FSW) together with the introduction of new aluminium alloys with improved weldability.

WEL-AIR focused on the manufacture of the stiffener and/or clip - skin connections using welding techniques: laser beam (Nd:YAG and CO2 technologies) and friction stir processes. Currently, the stiffener is welded on the skin in simultaneously conditions with two laser sources and filler wire. In the project, welding configuration with one-sided welding has been also developed. As competitive concept and technology, friction stir welding technique was developed for joining this stiffener-skin connection. For this alternative design, the joining was performed using an overlap of the stiffener on the skin. Short distance weld configuration required the use of the retractable pin tool technique.

The weldability of the new candidates using laser beam process has been investigated through three different approaches. First one aimed at developing and using an analytical approach based on the RDG (Rappaz- Drezet-Gremaud) hot tearing criterion in steady state condition. This analytical tool has been developed by the Computational Materials Laboratory (LSMX, EPF-Lausanne), subcontractor of EADS Innovation Works. It allowed to classify the WEL-AIR alloys according to their hot cracking susceptibility (HCS) without filler wire, to study the influence of the filler material and its dilution on their HCS and finally to investigate the potential of hybrid welding cases with dissimilar material joining.

The main achievement of this new stringer end design is the elimination of overlapping of both welding starts (or ends) and highest stress concentration locations. Hence, circular shape of the stringer ends and continuous welding process aim to prevent possible solidification cracking and later easy initiation of the fatigue crack.

The design suggests using two welding processes:
1. current double-sided LBW for welding of stringers by starting and finishing of the welding process at the beginning of the curvature;
2. use of Nd:YAG process with robot arm to complete the welding of stringer ends as one-sided welding by starting and finishing of the welding at the stringer sites away from stringer ends.
The finite element analysis established the degree of stress relaxation at the curved or ringed stringer end with the new GKSS approach.

Use of sensors for stringer start and stop detection and ramping of laser power at the beginning and end of the welding process was also recommended for defect free welding of stringer start and stop ends. This allowed accurate tuning of welding parameters: laser power ramping down from the power for continuous fillet welding at the stringer end. For avoidance of crack formation within the run-outs of the fillet welds at the stringer ends it was also necessary to use welding filler material with 5 weight percent silicon, to guarantee a minimum value of more than 3 weight percent within the melt zone.

According to the different technological maturity of the LBW and the FSW T-joint concept, concept validation step has been differently conducted for both technologies. LBW concept has been limited to the manufacture of four stringer panels implementing LBW of a 90 degrees cut stringer with an end-hole welded using micro-extrusion filler wire (EADS Innovation Works) or three additional welded clips between stringers (GKSS). Residual strength test conducted after fatigue crack growth exhibited comparable behaviour of the GKSS 4-stringer 3-clip panels with the base metal and the 4-stringer panel.

The friction stir welding has been performed on 1,6 mm thick L shaped stringer and skin element. Both similar configurations using the 2024 T3 and 2139 T3/T8 aluminium alloys and the dissimilar configuration defined as a PA765 L shaped stringer on a 2139 skin have been developed. For these three configurations, process parameters for short distance welding have been selected using retractable pin tool technique. Three resulting hardness profiles are examined: one at the mid-thickness of the stringer, one crossing the weld nugget and the last one between the skin lower side and the weld root. These hardness distributions show that whatever the alloy configuration the lowest hardness point of the joint is located in the skin and the weld nugget.

Corrosion behaviour of the FSW stringer-skin concept has been investigated through various testing conditions, accelerated intergranular corrosion sensitivity test, marine natural exposure, salt spray including the effect of protection routes. All the tests showed that the upper side of the weld is sensitive to corrosion but the degree is very dependant to the alloy and the temper.

Damage tolerance specificity of the run-in and run-out location has been investigated on welded clip specimens by GKSS. Fatigue crack growth behaviour has been studied and analysed for different welding sequences and the presence or no of pocketing. Behaviour analysis required the determination of the residual stress field using the neutron scattering method. One run-in / two run-out configurations is highlighted as the best welding routes regarding the residual stress type (i.e. compressive transverse residual stresses are generated at the second run-out) and crack growth rate.

For FSW technology, three types of welded technological specimens have been manufactured. First one was dedicated to assess the behaviour under typical bi-axial testing by DLR. Single stringer and stringer-frame panels have been manufactured by EADS Innovation Works associating various alloy combinations. The last both specimens were directly linked to the demonstrative to P180 Piaggio Aero.

Industries twin turboprop business aircraft fuselage barrel. Before dealing the clamping tool design of this barrel, major principles of clamping and welding including the welding procedures resulted from the EADS Innovation Works technological transfer have been explored through the manufacture of 2-frame specimens and butt joining specimens (Institut de Soudure). From these preliminary investigations, clamping design of the P180 fuselage barrel has been set-up by Dassault Aviation and Institut de Soudure delivered two barrels with one dedicated to a pressurisation test.

WEL-AIR provided new knowledge related to short distance welding and its related behaviour. Short distance welding using laser beam welding technology or friction stir welding is then viable concept and it requires to consider design aspects, welding parameters but also welding equipment reliability and monitoring accuracy.

The weldability of the WELAIR alloys has been demonstrated but the savings they offer on their unwelded conditions are not saved after welding due to the local weld property. Further investigations including new weld pool composition would be needed. The FSW T-joint concept with overlap has been confirmed as possible technological solution for the manufacture of the stiffener-skin connection.

Related documents