3DPrintingTechnology:FromNumericalSimulationtoApplications
ConferencePaper·August2015
READS
165
6authors,including:
H.Nguyen-Xuan
HoChiMinhCityUniversityofTechnology(…
150PUBLICATIONS3,640CITATIONS
SEEPROFILE
ChauNguyenKhai
Vietnamese-GermanUniversity
3PUBLICATIONS0CITATIONS
SEEPROFILE
TranHuynh
HoChiMinhCityUniversityofTechnology(…
6PUBLICATIONS22CITATIONS
SEEPROFILE
Availablefrom:ChauNguyenKhai
Retrievedon:25April2016
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
3D Printing Technology: From Numerical Simulation to
Applications
Nguyen Xuan Hung1,2, Chau Nguyen Khai1, Luong Tan Vu3, Tran Huynh
Trung Nam1
1
Center for Interdisciplinary Research in Technology (CIRTECH), Ho Chi Minh
University of Technology (HUTECH)
2
Engineering and Numerical Simulation Company (ENSCO)
3
Mechatronics and Sensor Systems Technology, Vietnamese-German university
Corresponding author: ngx.hung@hutech.edu.vn
Abstract
In recent years, 3D printer has become an efficient auxiliary device for manufacturing in experiment as well as commerce. The preeminent features of 3D printing are to provide a close connection of numerical simulation and design for actual structures. Current products contain complicated geometries which are still challenging with traditional methods can be now created promptly and accurately by this technology. Especially in heavy industry field, the prototype scale models are helpful to make preliminary design before fabrication. In this paper, the combination of 3D printing technology and finite element analysis is introduced as an ideal connection from numerical simulation and design procedures to commercial products, which have been preformed successfully at our company.
Keywords: 3D Printing, ANSYS, Catia, Finite element analysis 1. Introduction
First of all, simulation or visual experiment is crucially an essential task for almost engineering problems. In the cases that analytical or exact solutions cannot be found easily, approximate solutions have been devised. It is also a replacement of actual experiment because of high cost and time–consuming. The combination between simulation and numerical tools can offer an optimal way for problems in practice. It hence closes the gap between imagination and reality eventually.
Along with the fast development of computer science and innovations of computing algorithms, the numerical simulation (NS) supplied a global solution and played an important role in manufacturing process. As a result, a series of new products are always created promptly because the designs can be updated and validated in computer systems directly. Ameliorated technology has helped people to consider the internal structure of objects more conveniently. As the design cost might increase in direct proportion to the number of iterations for sample fabrication, developing many intermediate geometric models of the design and then inspecting the visually simulated
1
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
model before fabricating the physical model might help avoid unnecessary part fabrication for the design verification purpose.
In other important aspect, optimized structures can generally be defined as a process which includes determining the optimal value of design variables for technical designs. An objective function or criterion is maximized or minimized while requiring a set of geometrical or behavioral constraints. There are three types of optimization problems in Computer Aided Engineering (CAE): sizing, shape and topological optimization. Each of these has a wide field of research from approach to building computational models and algorithms, etc. For example, in the aerospace and the automotive industries, the topology optimization has become very popular. Two following types of the optimization problem are widely applied in many fields of civil, other designs and manufactures, cf. Figure 1.
Size optimization considers a structure which can be separated into a finite number of components. Each component is parameterized so that only one variable define it. Size optimization will try to find the optimal value of the parameter to match the requirements of the problem. Shape optimization is an extension of size optimization in that it allows more freedom of connections between components within the structure. The design allowed is limited to a fixed topology. The number of optimal variables can be limited but more than the size optimization. This is an improvement. Topological optimization is a mathematical approach to finding the optimal material distribution for a given designs domain. It gives no limitations to the structure that is to be optimized. In structural problems, topology optimization includes determination the position, shape and numbers of holes and type of connection in designs domain.
Figure 1. a) Sizing, b) shape and c) topological optimization [3].
Although NS can find optimal mechanisms to satisfy with engineering standards, it is usually not easy to adulterate. The improved designs also need technological breakthroughs in producing. Therefore, we need a new piecemeal technology to reduce the time and cost in molding operations.
Three dimension printing technology (3DPT) is an advanced fabricating technology based on accumulation knitting method, or addictive process, in which an object is
2
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
created by laying down successive layers of material until the entire object is built. Each of such layers can be seen as a thinly sliced horizontal cross-section of the printed object [4].
After simulating a model, people want to make a visual product in order to test the characteristics or functions. Moreover, development of intermediate geometric model and inspection of the visual model are very important as they help prevent the unnecessary parts of fabrication. However, the cost of traditional techniques can be quite expensive for single model and it takes time from deliver simulating file to factory.
In this paper, we introduce a new concept from designing to manufacturing. This is a combination of numerical simulation and 3D printing, which brings the ideas closer to conventional manufacturing methods and applications. 2. On Numerical Simulation to 3D printing
The numerical simulation procedure is summarized in Figure 2. At first, the geometry is drawn by CAD software. This geometry is similar to the considered real structure. Some of details can be eliminated to make the geometry simpler. Usually, the expected results should be determined at the beginning. We must decide what parameters are not influence the results to get the correct geometry. The geometry should be simple but it must satisfy practical problem.
In the second step, properties of material must be specified rely on the feature of the problems. Almost the factor is determined based on experiment in many times. Therefore, the simulation result is definitely subordinate these experimental parameters. Grid generating implements a finite element discretization by dividing the 3D structure into solid elements. In this period, the physical geometry is converted to finite geometry using meshing algorithms. The total number of elements will affect the accuracy of the simulation results and the demanded computer resources. A finer mesh with more elements leads to more equations which need more time for its assembling matrix and solving process. In addition, the structured mesh should be created to reduce the total elements and increase the convergent rate. The mesh generation is usually performed with commercial meshing tools.
Boundary conditions define fully constrains of the simulation model. There are two kinds of boundary conditions in the computing model. Essential boundary conditions are conditions that are imposed evidently on the solution and natural boundary conditions are those that will be satisfied by solver automatically. In the case of Finite Element approximations, the essential conditions will be exactly satisfied but the natural conditions only up to the order of the method. In many cases, the essential conditions correspond to Dirichlet' boundary conditions when the problem is written as a boundary value problem for a partial differential equation. The natural condition corresponds to a Neumann condition, a stress–free condition, or something similar, depending on the problem.
3
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Verification is the process which compares its implementation to the conceptual model. The model is tested to recognize and fix errors in execution. The purpose of model verification is to guarantees that the implementation of the model is correct. Validation checks the validity of the model's representation of the practical system. Model validation is the computerized model representing the real world as observed in the experiment.
Geometry modelling
Material modelling
Grid generating
Verification & ValidationSolve & post processing
Apply BCs
Figure 2. Numerical simulation procedure.
A designed model in STL file is transfer to 3D printer software in order to make instructions for 3D printer. Then those instructions are downloaded into the printer via USB cable or SD card. Finally, the filament will be melted and squeezed out onto the build plate. The object will be built layer by layer in thin lines. This method is called fused deposition modeling [1]. Work flow given in Figure 3 shows how to print a model with 3D printing. There are several important steps: Step 01: Choosing a model
STL (STereo Lithography) is a format file for 3D system, which are often used in rapid prototyping and computer–aided manufacturing. There are many modeling programs that could be used to design and create STL files which have become the standard file type for 3D printable models. For example, CATIA ®, SolidWorks ®, Rhinoceros®, and most Autodesk® programs, SketchUp®. Step 02: Preparing a model Filament
There are many type of filaments with different materials and colors. ABS (Acrylonitrile Butadiene Styrene) and PLA (Polylactic Acid) are popular filaments used in 3D printing. PLA is a natural, non–toxic and decomposable material. ABS is made from a combination of three types of plastic. These types of plastics can be mixed in different proportions to formulate particular ABS. Due to its flexibility, ABS is consumed mostly for mechanical designs or those that have interlocking or pin-connected pieces.
4
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Choosing a modelPreparing a modelPrinting and checking errorsSTL fileFilament typeDesign a modelResolutionRaftSupportAdvanced
Figure 3. Work flow how to print a model with 3D printing.
Table 1. Property of filaments [2] FILAMENT Glass Temp Melting Temp PLA 60~65° C 150~160° C ABS 105° C Amorphous, no 'true' melting point 230° C Nozzle Temp
Resolution
215° C This parameter can define the thickness of filament following the extruder with the value about 0.1mm - 0.3mm (ideally).
5
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Raft
This function can be chosen if the object is intended to be printed with a raft. The raft acts as a base and support structures for the object, it ensures the adherence of the build plate. The raft can be easily removed once on removal of finished object from the build plate. Supports
This function can be chosen if the object is intended to be printed with supports. The printer automatically generates supports for any overhanging sections of the object. Supports will be easily removable once you remove your finished object from the build plate. Advances
This function can be used to adjust additional options, including quality, temperature, speed and object strength. Specific elements that this function can adapt are quality, infill, shells, layer height, temperature, extruders, build plate, speed, speed while extruding and speed while travelling:
- Quality: each quality option affects particular section of the printed object, such as the strength and finishing.
- Infill provides the object with a varied solidity internal structure.
- Shells are the outline layers of the printed object. Every printed object needs at least one shell. Additional shells will be printed parallel with the previous shell to increase the object’s strength.
- Layer height determines the thickness of each printed layer. Thinner layers result in smoother surfer but longer printing time.
- Temperature helps control temperature of the build plate and the extruders. The necessity of this function depends on the material.
- Extruder needs to be heated to approximately 230°C before printing. This temperature is closely related to the extrusion speed and the material being extruded.
- Build plate helps printed plastic objects stick to have the printing platform without corrupting. Step 03: Printing and checking errors Objects gliding on the build plate
If printed objects does not stick to the build plate, the following solutions are required:
- Re–level the build plate: inconsistent plate height will lead to inconsistent adhesion.
- Ensure that the plate is clean: bubbles, scratches, dust, and oil from hands can prevent objects from sticking to the Kapton tape.
- Increase the temperature of the build plate by five degrees.
- Loosen each of the plate–leveling knobs about a quarter of a turn to bring the plate slightly closer to the nozzles.
6
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Warping and cracking
If the printed object starts curling on the build plate, or the new layer coming from the extruder starts shrinking and causing residual strain on the previous layer as Figure 4, try the following solutions:
- Increase the surrounding heat of the object in case printing without an enclosed printer.
- Make sure the closure is closed to keep a constant temperature during printing process.
Figure 4. Warping and cracking.
Figure 5 shows a flow diagram in manufacturing improvements using 3DPT. The cycle in the chart describe how to find out an optimal design through many iterations. This final design will be exported to CAD file and imported to printing control software of 3D printer.
Depending on demand of users; the output product can be used as an ideal prototype or a commercial manufacture. 3. Applications
3.1. Topological optimization for bridge structure
In the ideal conditions, the bridge structure deformed under uniform distributed load and has fixed support at two ends. We want to obtain the best structure corresponding to these boundary conditions and loading. Due to the symmetry of loading and working conditions, we just considered a quarter–reduced model to save the time–computing because we might have many iterations in heuristic updating scheme. The rectangular parallelepiped design domain and its behavior are shown in Figure 6. The “yellow” part in Figure 6a is fixed domain and it never change after optimizing. It means we just modify the material density in the “grey” part in the optimization loops to get the appropriate structure. The equivalent stress field showed the maximum stress value at the fixed end in Figure 6c.
7
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Figure 5. Flow chart of process.
(a) (b) (c)
Figure 6. The initial domain in design and under loading.
In
Figure 7a, the topology of initial geometry has been modified by changing the material density of every element after several optimization loops with respect to the volume constrain is 0.25. This result illustrated the connections between the bridge foundation and its supporters. However, some regions of these parts quite rough due to initial grid model. The final design in
8
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
Figure 7b can be achieved after smoothing and expanding numerical result. All smoothed surfaces are created through approximation at boundary elements.
(a) (b) (c)
Figure 7. From optimal result to practical prototype.
This is an arch bridge structure with abutments as architectural curved arches at each end. It works by transferring the weight of the bridge and its loads partially into a horizontal thrust restrained by the abutments at either side. The real prototype in Figure 7c can be constructed by 3D printer. For the convenience, extrude direction is chosen from top side to bottom side to save material in support printing. 3.2. The multiplier fan
The multiplier fan is also called as “bladeless fan” because it has not any external blades. It blows air from a ring with blades are hidden in its base. Dyson’s company has developed its designs from first idea created by Toshiba. By accelerating air over a profile like airfoil, it eliminates the buffeting and turbulence associated with these traditional household fans. Initially, Dyson engineers relied on physical prototyping for design development. Then they used CFD software from ANSYS to complement experimental testing and reduce development time [5]. A thin, high–speed sheet of air that induces surrounding air through a fan is produced. This unique provided a much smoother movement of air without external blades. Basically, the impeller is put inside the base to draw air from outside and lead it to the ring. The air is accelerated through an annular chink and passed over an airfoil–shaped ramp. The overall process takes a total of two weeks including building ring design, measuring, and hand finishing, assembling and testing the ring. Therefore, it also takes a lot of time and cost when the ring is redesigned.
The initial prototype is first considered in 3D problem as in Figure 8a, steady–state, incompressible, turbulent air flow using the k–epsilon turbulence model. A series of design iterations are evaluated with the main goal of increasing the amplification ratio. That means the fan must to move the possible maximum amount of air for a given dimension and power consumption. There are some factors having major effect on performance: the crack in the ring, the profile of the ring and the external ramp. The numerical results in Figure 8b show the pressure distributions at the sections in front of and behind the fan. The streamlines in the central plane of computing domain demonstrate the flow is relatively laminar in the far–field of the fan. For the convenient in assembly the impeller when the fan is actuated, geometry of the fan can be
9
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
subdivided into two parts (base and ring). These parts are fabricated separately as in Figure 8c.
3D design (b) Flow simulation with ANSYS (c)Test with small generator
Figure 8. From design to fabrication.
3.3 Pressing machine
The hydraulic pressing machine contents the main frame which is around 8 ton and hydraulic cylinder which pressing force 260 ton by 210 bar pressure. Using Catia to design and propose the solution of space as customer requirement so that operator can work easily and safety. The next step would be strength check of material of main frame and welding at connection according to Eurocode 3 standard by FEA in order to improve the safety factor and decrease the weight of machine. Due to its heavyweight, prototype of (1:300) scale model is printed by 3D technology to have an ideal for operating and connecting the machine plates in workshop (see Figure 9). Furthermore, this scale model has more advantage to obtain the solution of lifting and moving before processing to manufacture the real machine. Also, there are more popular applications such as gear, jack screw, light support are widely used in mechanical field. 4. Conclusions and future work
3D printing is a robust connection between simulation and commercial manufacture by constructing promptly layer-by-layer complex design. It is also a good alternative option to reduce manufacturing time and prototyping cost. It brings artistic design into real life and makes to increase residual value of industrial products. Some points are drawn as
- The 3D printing provides an equipment for special experiments in CFD, mechanics, electricity, etc.
- It saves the space for laboratory; more affordable, single plastic objects.
However, risk and challenge consist limited size of products, metal products, mass production. In many cases, it takes much duration in making the support. Herein, the printing time and material can be saved by optimizing the supports.
10
Hội nghị Khoa học toàn quốc Cơ học toàn quốc lần thứ XII
Thành phố Đà Nẵng, 3-5/08/2015
By the way, 3D printing technology has opened more great opportunity for practical applications and many research subjects are promising for future work.
(a) 3D design and manufacturing drawing with CATIA (b) FEA to check frame and welds
(c) 1:300 scaled model printing
(d) process manufacturing
Figure 9. Application of modern design with 3D design FEA, 3D printing and
manufacturing. References
[1] MakerBot Replicator2X User Manual
[2] http://store.makerbot.com/filament access on July 3nd, 2015
[3] M. Bendsøe, O. Sigmund (2003). Topology optimization. Theory, methods and
applications. Springer, Berlin. [4] Jee, H. J., and E. Sachs. \"A visual simulation technique for 3D printing\" Advances
in Engineering Software 31.2 (2000): 97-106. [5] ANSYS ADVANTAGETM Excellence in Engineering Simulation Volume IV:
Issue 2 (2010).
11
因篇幅问题不能全部显示，请点此查看更多更全内容