By Andrew Albritton
As seen in LIA TODAY
LASER ADDITIVE MANUFACTURING CHALLENGES
Laser Additive Manufacturing (LAM), as it pertains to powder based manufacturing, is a technique that utilizes the interaction of lasers and base materials to construct a product, rather than removing material from a pre-constructed block of material. LAM is quickly approaching the critical point of being more than a method to produce prototypes and small runs of one-off parts – it is poised to turn everything we know about mass production on its head. Professor Dr. Minlin Zhong, President-Elect of LIA and Director of the Laser Materials Processing Research Center at Tsinghua University, believes it surpasses all available alternative methods. Prof. Dr. Zhong says “LAM shows obvious advantages on freeform manufacturing, including free geometry, free structures, free strengthening mechanism, free microstructures, free performance and even free scale (from macro, to meso, to micro, to nano),”. Manufacturers who use LAM are able to reduce the waste of materials commonly associated with traditional subtractive manufacturing methods; decrease the weight of parts by cutting out filler materials; and have more control over material properties resulting in stronger, more complex, lighter, and more efficient parts. With such exceptional technology currently at our disposal, why hasn’t LAM been more widely adopted?
One of the most commonly cited reasons is that the costs to produce parts through LAM are prohibitive. The key driver of these high costs is that the supply chain for metal powders is not yet optimized for LAM technology. Materials are expensive, custom made, or not readily available. The Metal Powder Industries Federation (MPIF) states in its 2017 PM Industry Roadmap that, “A better understanding of the precursor materials impact on the metal AM process is required. Traditionally, precursor materials have been existing thermal spray powders that have not been refined/tuned to the AM process limiting optimization.” LAM parts producers are often using metal powders that have not been designed for use in LAM processes, which frequently results in suboptimal products.
According to MPIF, as of 2017, there are approximately 12 suppliers of metals for Additive Manufacturing (AM) for the international market, most produce stainless steel, cobalt-chrome, and titanium, with a few supplying aluminum alloys, copper, super alloys, platinum, Inconel, tungsten, molybdenum, and tool steels. With so few suppliers and a sparse number of common material types, there is a bottleneck for providing quality affordable metal powders to the LAM industry. With companies expanding the selection of materials that can be laser processed, it is vital that the problem of material availability be resolved. For example, Nuburu has produced a “blue” laser which operates at the 450 nm wavelength, and is capable of processing gold, aluminum, brass, and copper.
SUPPLY AND DEMAND
What can be done to improve the supply chain and reduce the cost of LAM part production? The metal powder industry does not supply enough quality powder to support widespread adoption of LAM, while early adopters of LAM applications do not create enough demand to drive competition into the metal powder market to reduce prices. A first step to get these industries operating in unison will be the creation and mass adoption of standards, specifications, and best practices in regards to metal powders. By standardizing metal powder properties for best final product properties, metal powder suppliers would be able to build up an inventory without relying on custom special orders. Specifications on how surplus powder from a project can be reused could also help introduce addition cost savings to manufacturers.
STANDARDS FOR QUALITY CONTROL
Another hurdle for LAM is microstructural quality, uniformity, and repeatability. To become a replacement for more legacy manufacturing methods, LAM needs to produce parts consistently and continuously that are to specifications. With traditional subtractive manufacturing methods, there are several quality control points where product is inspected and defects are addressed prior to the next step, resulting in no wasted effort past the point of failure. With LAM, the part in question is created from the ground up; this determines the final product’s quality, microstructure, and mechanical properties simultaneously. The process is completed with either a perfect or defective final product. Paul Denney, Director of Advanced Process Development with IPG Photonics, states, “Unlike machining where you start with a “block” of material with known quality and properties, additive production of parts requires a combination of motion with the prediction of the microstructures, mechanical properties, and stresses. Because the properties are closely connected to how the material is deposited, this greatly complicates the development of processing procedures and parameters.”
What methods can be implemented into a given LAM process to help ensure quality of the final product? The first quality control concerns are addressed long before the process begins. Starting materials must be certified as appropriate for the application, the order of operations of the production device should be scrutinized to ensure that the final product will be to spec with minimal waste, and the machine itself must be operating at peak parameters. As the production of a LAM product can take an extended amount of time, any loss of power to the point of interaction can have detrimental effects to the end product and even the products in queue. Loss of power can be caused by an actual power failure, a dirty or damaged optic, or other origins. With the structural integrity of a LAM part resting critically on the success of every step of the process, it is imperative that the process is stringently optimized and the machine is operating at peak performance. Here is what Paul Denney has to say about the subject:
“Because of the additive manufacturing approach in bed based systems, even if defects can be detected and possibly ‘corrected’, any changes may not be possible. An example of this may be what is done if a ‘defect’ is flagged in a single part in a batch of parts being produced. One approach would be to stop the processing and ‘correct’ the defect. However, if this is done then the thermal history for all of the parts may be altered and all parts may now be out of the desired properties. Another approach would be to stop processing on the part with the defect, but this again would alter the heat load on the complete batch or the time between other parts being produced which may again alter the properties. So any monitoring system will need to detect changes prior to the formation of any defects while at the same time any corrections must be made within the acceptable parameter range.”
There is a thin line between success and failure: one small interruption can ruin an entire batch of product. What can be done to prevent this?
As Paul explained, this is not a single issue, LAM processes need both a method to detect defects and the ability to immediately respond to them. A starting point is to ensure that redundancies are incorporated into the build process so that if a common defect occurs at a certain stage, there are defined responses the system can take automatically to correct them. In the case of a laser lens issue, it may be beneficial to incorporate additional laser delivery systems to the process as a redundancy to pick up where a suboptimal device has failed in real time.
EVALUATING THE FINAL PRODUCT
In addition to inline defect detection, the industry as a whole will require a standardized best practice for evaluating finalized parts. For traditional manufacturing methods, a sample of the produced part pool is selected for evaluation via destructive and non-destructive tests to certify whether a set of parts are built to specifications. As many LAM-produced parts are complex and costly to produce, it seems wasteful to destroy a set of them to certify them. In the paper “Evaluation of 3d-Printed Parts by Means of High-Performance Computer Tomography” presented at ICALEO 2017, authors Lopez, Felgueiras, Grunert, Brückner, Riede, Seidel, Marquardt, Leyens, and Beyer reviewed the viability of X-ray Computer Tomography (CT) and 3d scanning as methods to detect inferior AM parts. The paper concludes that the CT method best fits the needs of the AM industry. According to Lopez et. al, “Computer tomography can quantify all complex structures in scope of the proposed demonstrator and delivered deviation values of the measured structure, providing a good base for comparison across demonstrators made by different methods, materials and dimensions. Porosity or defects down to 3 µm can be determined by the used CT system.” Currently, CT scanning a LAM part is a time consuming process, but with additional focus on improvement it could become an essential quality non-destructive control method for finalized parts to evaluate complex internal structures.
TOO MANY ALTERNATIVES
A third barrier to the spread of LAM is the multitude of alternative methods in the industry. As stated by Prof. Dr. Zhong, “Some conventional metal deposition technologies such as arc building-up welding, plasma building-up welding and electronic building-up welding can also fabricate metallic components in near shape. Their deposition rate and productivity may be high and the costs may be lower, but normally they are limited in fabricating complex geometry and accuracy.” Freeform manufacturing is where LAM excels, but despite its many advantages over alternative methods, it has an Achilles heel.
One advantage of alternative manufacturing methods is the speed at which a product can be produced. However, according to Paul Denney, this speed gap is closing faster every day.
“While higher laser powers allow for higher deposition rates but at the expense of lower resolution, some researchers are looking to maintain the resolution by combining multiple lasers into an additive deposition system. Research groups and equipment builders are investigating how best to handle multiple lasers in the same processing area. There are other areas that may be investigated including power distribution to improve the interaction between the power and laser beam to improve efficiency of the process and to minimize defects. This could improve the deposition rates while at the same time maintaining quality.”
Prof. Dr. Zhong hopes that soon LAM researchers will, “improve the materials diversity, increase the dimension (to square meters), increase the deposition rate and decrease costs. A hybrid approach to combine LAM with the conventional additive manufacturing methods may be a solution to achieve the above targets.” The concept of a hybrid production system that can combine multiple lasers with fast alternative methods where complexity is not a requirement could lend itself to faster build times.
THE LATE ADOPTERS
Earlier in the article, we touched on the final barrier to the wide spread success of LAM: industry standards. Current standard offerings from ASTM and ISO cover Design, Materials and Processes, Terminology, and Test Methods. Additionally, new processes are created frequently and new standards are being developed every year in an attempt to keep up. It is unclear how much of the industry has adopted these existing specifications. Until the entire market accepts a set of standards for all steps of the Additive Manufacturing process and supply chain, the evaluation of AM parts will remain a costly endeavor that will limit AM’s potential. MPIF expresses a bleak outlook on metal AM in its State of the PM Industry in North America – 2017 document: “Despite all the fanfare, true commercial long-run production still revolves around only three product classes: titanium medical implants, cobalt-chrome dental copings, and cobalt-chrome aircraft nozzles.” The truth of the matter remains that without a set of clearly defined standards, the LAM industry will continue to remain confined to early adopters like the Aerospace and Medical fields. With the benefits in intricacy and weight saving advantages LAM should have obvious opportunities in the automotive and electronics industries.
Markets are watching LAM for innovative uses before taking the plunge and embracing the technology. Currently, LAM may appear to have a bad Return on Investment (ROI) if producers only hope to replicate their existing products through LAM rather than innovating their parts to capitalize on its strengths. In the words of Paul Denney, “If AM is supposed to make big impact, companies are going to have to rethink their parts; determine how AM allows for changes in the design and possibly improve the performance. The benefits can come in many forms which could be a weight savings, a production savings, and/or a performance savings.” The industry needs to challenge its way of thinking about production to allow the benefits inherent to LAM to propel their production and parts to new levels of performance. Paul Denney provided the following illustration: “With the formation of properties ‘locally’ instead of in ‘bulk,’ it is possible to produce ‘gradient’ materials. The ‘gradient’ can come by changes to the properties of a given chemistry of material or by using materials with different chemistries. As an example: a bracket could be produced for a jet engine that has high temperature properties near the engine but as the bracket extends to an attachment point, the properties/chemistry can be altered to improve the fatigue properties.”
LAM has a bright future and many engineers and scientists are working to unlock its full potential. Once the barriers of the supply chain, dynamic quality control, speed of production, and process standardization have been resolved, it is highly likely the LAM will be a manufacturing method of choice.
Paul Denney, Director of Advanced Process Development with IPG Photonics and LIA’s Past President
Prof. Dr. Minlin Zhong, Director of Laser Materials Processing Research Center at Tsinghua University
and LIA’s President-Elect
Lopez, E., Felgueiras, T., Grunert, C., Brückner, F., Riede, M., Seidel, A., Marquardt, A., Leyens, C., Beyer, E. (2018). Evaluation of 3D-printed parts by means of high-performance computer tomography. Journal of Laser Applications 30, 032307; https://doi.org/10.2351/1.5040644