Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Laser Metal Deposition for Large Scale Net Shape Manufacturing

Project Code: 30153

Objectives

  • Identify two industrially relevant high alloy materials (e.g. Haynes 230, Hastelloy X, Inconel 625 or Titanium) suitable for net shape LMD manufacture of annular geometries with a wall thickness in range 0.8mm-1.6mm. Assessment metrics will include melt pool stability and wall distortion. This will be supported by modelling of thermal behaviour to help ring-fence material choices
  • To model transient thermo-mechanical behaviour during the manufacture of thin walled axisymmetric structures in order to assess and predict distortion patterns (support of Task 1) and support the implementation of corrective measures in the toolpath (support of Task 5)
  • Benchmark net-shape build capability of a commercial (unmodified) coaxial nozzle against TWI modified nozzle using the combustor casing geometry (possible generation of IP)
  • Develop plugins to the ToolClad software to generate relevant toolpaths in support of demonstration activity
  • Manufacture two new demonstrator components, one in Inconel 718 and one using a material identified in Task 1, in size range 400mm to 750mm diameter for generating business through conferences, client visits and publications. For example: 
  1. A closed volume demonstrator (e.g. dome ended vessel)
  2. A non- axisymmetric demonstrator (e.g. engine cowling) 

Project Outline

Laser metal deposition involves melting a powder filler material using a laser beam, forming a layer on a base material with a metallurgical bond. Using multi-layering techniques and a modified coaxial nozzle, TWI has recently developed a refined version of the process that creates thin walled cylindrical structures to net shape with good surface finish. The CAM software which enabled this work was developed in an internally funded project. From this work, an Inconel 718 helicopter combustor casing (90mm tall, 300mm diameter, 0.8mm wall thickness and weighing 0.7Kg) was successfully manufactured to dimensional tolerance (herein called combustor casing). This CRP seeks to broaden a technical competency, attract more industrial interest through the creation of further industrially relevant demonstrators and increase the impact of a world leading capability.

The current state of art for net shape manufacturing of 3D components using LMD most probably resides with TWI, both in the software it has created (ToolClad CAM software) and the LMD process parameters it has developed.

For the software, CAM style tools numerically and adaptively slice a CAD data file (STL format) into helical layers. A five axis toolpath and LMD deposition parameters are then mapped onto the slice data to guide a three-axis coaxial LMD nozzle across a moving substrate manipulated by a two-axis CNC rotary table. With precise synchronisation of the movements of rotation and tilt of the substrate with small incremental movements of the coaxial nozzle, a continuous spiralling weld track can be deposited or ‘grown’, layer on layer, out of the substrate. This helical multi-layering technique has been successfully demonstrated for the manufacture of components that require a constant rotational speed, a fixed substrate tilt and a fixed radial position of the nozzle for any given layer of rotation (i.e. axisymmetric components). For non-axisymmetric geometries, a more complex toolpath is required, analogous to a pen moving around a Spirograph.

For the deposition parameters, careful control of the powder-gas beam focus and precision of nozzle standoff  on top of a rising and re-orientating wall is critical in order to maintain a stable melt pool for several hours and for several Km of weld (8 hour build time for the combustor casing). If the melt pool becomes unstable, or begins to wander, then the wall density and surface finish will reduce and the probability of build failure heightens. Beam wander is also caused by the thermal expansion and contraction characteristics of different alloys, giving rise to different levels of wall distortion. A balance of all of these factors has been successfully found for Inconel 718 powder. Although, a small amount of wall distortion in the radial direction still persists in the first 10-15mm of vertical build height before the effect diminishes. This can be anticipated and corrected (but not yet comprehensively managed) through toolpath adjustment. It should also be noted that it is currently unclear whether the use of a TWI modified commercial coaxial nozzle is fundamental to recent success.

Competing 3D printing technology for net shape metal component manufacture is laser and electron beam melting of a powder bed. However, comparatively these technologies have limitations (benchmarked against combustor casing):

  1. Small build volume (<300mm in xy for many commercial systems)
  2. Large powder material requirement (~50kg compared to 1.2kg for LMD)
  3. Large surplus powder (~49Kg of unused powder compared to 500g post LMD).

 

Maintaining part quality with the use of recycled powder is a highly debated subject in additive manufacturing, with no clear conclusion. Hence, in a waste conscious environment, LMD would be a very desirable approach for the manufacture of components that exhibit a large minimum-perimeter bounding box but with a small material volume.

Relevant Industry Sectors

Industry Need

Considerable interest has been expressed regarding capability to manufacture components with specific attributes, including: (1) large scale (400-1200mm in diameter), (3) concentric walls, (4) addition of monolithic features on inwards/outwards facing surfaces (bosses, lugs etc.), (5) non axisymmetric contours and (6) different materials (including multiple materials). These requirements are all feasible, but further progress at this early stage of development requires considerable commitment; LMD deposition procedures have to be developed for new materials and new geometries and CAM software tools for different geometric primitives have to be written and beta tested. There is a significant industry pull for this technology that requires the combined efforts of predictive modelling and the net shape production of high-value, complex parts.

For more information please email:


contactus@twi.co.uk