3D printing is finding its way into just about every industrial and manufacturing sector, but its introduction in turbomachinery has been relatively slow. Due to the extremely high temperatures, enormous pressures, high rotational speeds and large parts involved, turbomachinery has turned out to be one of the most difficult application fields for 3D printing, also known as additive manufacturing or AM.
However, the technology has evolved to the point where it can now produce viable turbine and compressor parts. In fact, it can create structures that are more efficient, more intricate and longer lasting than those made by conventional manufacturing methods.
For example, Siemens' SGT-700/800 gas turbine burner fronts were traditionally manufactured with 13 different parts and required 18 separate welds. A redesign and manufacture by 3D printing resulted in nozzles that have one component and no welds. These burner nozzles allow co-firing at higher combustion temperatures thanks to an improved burner tip design.
These and many other components can be produced on demand, eliminating long lead times for foundry-produced parts. As a result, 3D printed components can be produced ten times faster than conventional means.
On the maintenance side, AM opens the door to easier and faster repairs. The technology has recently been used to reduce the amount of material that has to be removed during burner tip repairs from 120 mm to only 24 mm.
But this is only the beginning. New techniques and technologies are continuously being added to the AM arsenal.
It has become a recognized fact about AM that just about anything a designer can dream up can be printed, with only a few exceptions. However, traditional design methodologies such as finite element analysis and mechanical integrity calculations have to be augmented with more powerful tools.
The design of 3D printing parts includes detailed simulations that demand huge amounts of computer power - far more than that typically required by CAD and other traditional design tools.
The reason for this is that AM is capable of generating complex lattice structures as opposed to solid components. When employing lattices along with a high number of structural elements, the computing power needed for simulation is enormous. In fact, the capabilities of 3D printing demand top-of-the-line computer hardware backed by advanced software. That's why sophisticated lattice structures are only beginning to be employed today.
Lattice structures offer many advantages for turbomachinery. Instead of a thick wall which can be tough to cool, you end up with a lightweight structure which is full of empty spaces that facilitate air cooling. As heat transfer is far easier, parts can be subjected to higher temperatures without sustaining damage. This allows designers to push the envelope in terms of hot gas path temperatures and more easily integrate effective cooling channels while extending the life expectancy of equipment.
A lattice structure enables a more even temperature distribution, leading to lower stress levels in components even in harsh environments. This is particularly important in hot gas path parts such as the combustion liners, end caps, fuel nozzle assemblies, crossfire tubes and transition pieces.
Another advantage of lattices is their ability to dampen or even eliminate vibration. Designers can run simulations to see how different lattice structures impact various vibration frequencies. This is a better approach than having to take alternative measures to dampen vibration or figure out how to work around a distinctive vibration that could be potentially damaging to components. Eigen frequencies, for example, which can destroy equipment, can be avoided without the need for remedial damping actions.
The possibilities are endless. Components can either be created with a lattice structure composed of the same cell type and a uniform orientation, or can have a change in the orientation of the cells. The size of cells in certain areas of the component can be adjusted, or designers can combine different cell types into a complex lattice structure.
3D printing in plastic has been around in commercial usage since the 1980s. But in metal, it only became commercially available around 2005. This technology takes three-dimensional engineering design files and transforms them into fully functional and durable objects. Metals and plastics - and now even ceramics - can be employed.
In the case of metals, a metallic powder is laid down by the 3D printing machine and a laser is directed across the powder to melt it. Layer by layer, as the molten metal solidifies, the component is built inside the printer. This technique enables the manufacture of geometries that would not previously have been possible.
Lattice structures, for example, are easy to print if they follow certain design rules. The only real limitation is that overhanging structures or horizontal struts can't be printed. Instead, lattices are formed at no more than about a 45-degree angle.
Downward-facing surfaces can be created by being built up one layer at a time by the printer. But a 45-degree overhang is the maximum possible - i.e., instead of cells with struts that look like '+' signs, designs must use 'x' shapes.
Advanced software is also required to make 3D printing possible. Similar to computer numerical control (CNC), the software determines where the equipment goes and what it does. In the case of AM, the software directs the printing head and the deposition of material.
While stand-alone, third party 3D printing software is fairly advanced, it does not yet integrate well with CAD and modelling tools. Current designers are forced to conduct a number of workarounds in order to move data from one system to another. But that has limited workability. Different formats are present and don't translate well from one to another. When you export or import lattice models into other software tools, quality can suffer. The tolerance demands of modern turbomachinery architectures do not allow for errors in quality, particularly in the hot gas section.
Accordingly, Siemens has created a new software tool that makes it possible to work directly inside the CAD environment with lattice structures. Known as NX11, this software suite allows the designer to remain inside one software environment, interact with CAD and maintain the high level of quality demanded by 3D printing of turbomachinery.
Take the case of a combustor. One way to prevent the flame from causing damage further up in the fuel system is via the inclusion of a porous lattice structure that can dissipate heat and prevent the flame burning upwards into a piping system. This is a simpler and more effective solution to the flame stopper elements currently utilized.
It's important to understand the many ramifications of 3D printing. Firstly, it may have been used primarily in prototyping in decades passed. But today, it becomes a proven approach to manufacturing. Certainly, casting can and should be used to create larger or less complex high-volume parts. But casting itself limits component geometry. 3D printing enables far more complex geometries as well as a means of reducing the number of welds needed in components.
Additionally, AM techniques break down barriers that might limit the design or manufacturing of turbomachinery parts. Previous technologies confined engineers and designers in the way they had to think about products and how they went about the business of specifying the types of conventional casting, forging or plate materials they were made of. With 3D printing, new possibilities through mixing and designing new powders become readily apparent.
Further, new sets of tools are being integrated into NX11 to enable companies to design, simulate performance and manufacture even more advanced 3D printed parts. For example, a topology optimization tool allows a designer to reshape a design into more of an organic shape that can be lighter weight and stronger than conventionally designed parts. It achieves this by optimizing stress paths through its geometry.
Another recent advance is the ability to automate the process of creating conventional design inputs, materials and loads. The system creates a variety of possible shapes for a component and the designer can use these to run multiple simulations to determine which approach would be best. Known as the convergent modelling tool, the designer's job is made easier as he or she can mix and match various types of geometry, both precise and faceted. Once the simulations are finished, the designer can decide which direction to take the design.
Existing parts, too, can be scanned and the digital data fed into NX11. Simulations will reveal any existing design weaknesses and uncover adjustments to optimize that part or make it easier to manufacture.
These new design tools eliminate a whole lot of wasted time in design and prototyping, while taking turbomachinery design to a new level of performance. Harnessing the computing power of the latest processors, the designer sets up the objectives of the design and describes what he or she wants to achieve. That unleashes thousands of design iterations and opens the door to new design concepts.
What possibilities will 3D printing eventually uncover? No one can say with any accuracy. Certainly, the design and manufacture of turbomachinery components is in for something of a paradigm shift. But beyond that, some anticipate that a revolution in material science is quite likely.
After all, the latest tools may well permit designers to conduct simulations of material properties to target higher performance, greater durability at high temperature, greater ability to cope with rapid power plant cycling and other desirable outcomes. This may well lead to the formulation of materials and alloys that haven't been conceived of today.
Andreas Graichen is Group Manager for the Additive Manufacturing Centre of Competence at Siemens Power Service in Sweden