A Method for Generation of Random Lattice Structure for Additive

Manufacturing

Dinh Son Nguyen1

1
The University of Danang, University of Science and Technology, Danang, Vietnam
([email protected])

structure is a set of unit cells repeating systematically in

Abstract – Additive manufacturing technologies provide three directions of a design space. The unit cells in the
us new opportunities to fabricate a product with much more lattice structure have the same configuration. Thus, it is
complex geometry and architecture that conventional also called the uniform lattice structure (see Fig. 1a).
manufacturing ones such as turning, milling, grinding,
casting and molding do not have the ability to implement.
Hence, lattice structures are becoming more popular and
used extensively in the industrial product design to minimize
the usage of materials due to the additive manufacture
technologies. Lattice structures are formed by the cellular
unit of structure repeated regularly or irregularly in a design
area. However, the creation and generation of these
structures for additive manufacturing in the current CAD
environment have many difficulties. Therefore, the paper will (b) Conformal lattice
propose a new approach to design the lattice structure, (a) Non-conformal
lattice structrue structrue
especially the random one. This approach allows helping
product designers to design and generate lattice structure
randomly based on the Voronoi diagrams.

Keywords – Additive manufacturing, random lattice

structure, Voronoi cells.
(c) Random
I. INTRODUCTION lattice structrue
Additive Manufacturing as a scientific term indicating
layer manufacturing processes is developed based on
Rapid Prototyping technologies. Additive manufacturing is
a fabrication technology different from subtractive Fig. 1. Different configurations of the lattice structure.
manufacturing. It is defined as a process of building a part The part of the uniform structure out of the design zone
by stacking a very thin layer of materials up to other layers. will be removed from the original structure in order to fit it
The current additive manufacturing technologies are able in the design area of a product. Unit cells near to the
to be categorized into seven different technologies as boundary surfaces are not the same configuration to the
follows: initial unit cell structure. So, it is also called the non-
• VAT Photopolymerization (DLP, CDLP, SLA) conformal lattice structure. These unit cells do not have the
• Powder Bed Fusion (EBM, Multi Jet Fusion, same mechanical properties as others. As a result, it can
SLS/SLM/DMLS) make the mechanical properties and strength of the whole
• Material Extrusion (FDM) product to be decreased. So, the conformal lattice structure
• Material Jetting (DOD, NPJ) is developed to adapt to the complex boundary of the
• Binder Jetting design space.
• Direct Energy Deposition (LENS, EBAM)
The conformal lattice structures are also able to be
The main advantages of additive manufacturing
built by many cellular structures. But each cellular
technologies are abilities to fabricate extremely complex
structure in the conformal lattice structure is geometrically
geometries, to build topology structure and change the
different from the non-conformal one. The geometry of
material architecture inside products. As a result, lattice
each unit cell is transformed in order to adapt to the
structures can be manufactured quickly by AM
geometry of the design space (see Fig. 1b).
technologies.
The lattice structure includes many bars connecting The random lattice structures are totally different from
each other regularly or randomly. The lattice structure can the non-conformal and conformal one. It is built by
be divided into different types as non-conformal, randomly distributed bars in the design area. The radius
conformal and random one. The non-conformal lattice

978-1-7281-3804-6/19/$31.00 ©2019 IEEE1275 1275

Authorized licensed use limited to: VIT University. Downloaded on February 01,2022 at 15:42:48 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2019 IEEE IEEM

and the length of bars in these structures are also different The part out of the design space will be removed and the
(see Fig. 1c.). rest of the random lattice structure will be assembled in the
All lattice structures can be fabricated easily thanks to final designed product.
the appearance of additive manufacturing technologies. So,
1) Voronoi diagrams
it is becoming a useful application as its special
characteristics as high strength, lightweight and stiffness. A Voronoi diagram is to partition a plane into different
Thus, lattice structures were applicated in many different regions based on the distance between center points to
fields in the industry such as automotive engineering [3], points in a specific subset of the plane. The definition of
thermal industry [1, 2] and biomedical engineering [4]. It Voronoi diagrams is presented in [11]. Let S ⊂ \3 be a set
is very difficult to design and create a 3D geometric model of n ≥ 3 points in the space. For points p = (p1, p2, p3) and
of lattice structures in the current commercial CAD q = (q1, q2, q3), the Euclidean distance d(p,q) between the
software. However, the methods proposed by [5-7] aim to point p and q is calculated by equation (1).
design non-conformal and conformal lattice structure.
These methods allow helping to design and create a 3D d ( p, q ) = ( p1 − q1 ) 2 + ( p2 − q2 ) 2 − ( p3 − q3 ) 2 (1)
geometric model and integrate it into the design zone of the We denote by B(p,q) a bisector between two point sites p
product. There is some commercial CAD software that and q. The bisector B(p,q) is the set of points equidistant to
supports to design only the non-conformal lattice structure. p and q under the distance d(p,q). It is defined by the
There are some studies concerning the random lattice following expression.
structures and their application [8-10]. Meanwhile, the
creation of the random lattice structures in the CAD B( p, q ) = {r ∈ \ 3 | d ( p, r ) = d (q, r )} (2)
environment is less concerned. Therefore, the paper It separates the half-region containing point p from the
proposes a method based on the Voronoi diagrams. It half-region D(q, p) containing point q.
allows product designers to be able to generate a geometric
model of the random lattice structures automatically and D( p, q) = {r ∈ \3 | d ( p, r ) < d (q, r )} (3)
integrate it into the model of the product in the CAD Let VR( p, S ) be the Voronoi region of point p with respect
environment. to S. It is defined by the following expression.
II. METHODOLOGY VR ( p, S ) = 
q∈S , q ≠ p
D ( p, q ) (4)

The random lattice structure is a type of lattice Finally, the Voronoi diagram of S is defined by the
structure in which bars are randomly connected with each following expression.
other in the design space. Especially, the value of the radius V (S ) = * VR( p, S ) ∩ VR( p, S ) (5)
and length of each bar can be also changed randomly. The p , q∈S , q ≠ p

random lattice structure is really difficult to be built Where Voronoi region V(p,S) is the intersection of n-1
compared to the conformal and non-conformal lattice open half-regions containing the point p.
structures. In order to generate a 3D geometric model in A set of randomized points is used as a set of seed
the CAD environment, the paper presents an approach points to generate the Voronoi diagrams by using the
based on the Voronoi diagrams. The overview of the algorithm proposed by [12, 13]. The data of the Voronoi
proposed approach is given in Fig. 2. diagrams including the edge of the Voronoi cell and
coordinates of points are used in order to build the frame
model of structure.
2) A frame model of random lattice structures
In order to form the frame model of the structure in a
design space, two different approaches are proposed in the
paper. The first approach as shown in Fig. 3 includes the
following steps:
• Step 1: Creating a box space
Fig. 2. The proposed method to create the random lattice structure.

From a set of randomized points, the Voronoi The geometry of the creation zone of the random
polyhedrons are created by using the randomized points as lattice structure is often very complicated, so it is
the centers of Voronoi cells. The data of the Voronoi necessary to create a box space surrounding the
polyhedrons generated by the Voronoi tessellation design space. The box space aims to generate a set
algorithm include coordinates of points and edges of of randomized points easier.
Voronoi cells. The frame model of this structure including • Step 2: Generating a set of randomized points
points and lines is built by using the data of Voronoi cells.
The 3D geometric model of this structure is established by A set of randomized points is generated uniformly
using the frame model and the value of the radius of bars. inside the box space. The density distribution of

978-1-7281-3804-6/19/$31.00 ©2019 IEEE1276 1276

Authorized licensed use limited to: VIT University. Downloaded on February 01,2022 at 15:42:48 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2019 IEEE IEEM

bars in the random lattice structure depends on the • Step 2: Generating a set of randomized points
number of randomized points.
A set of randomized points is generated randomly
inside each box space. Obviously, the density
distribution of bars in the random lattice structure
depends on the set of the randomized points. The
number of randomized points NP is calculated by
equation (6).
V
N P = DS (6)
S

Where VDS is the volume of the design space and S

is the average size of Voronoi cells.
• Step 3: Creating Voronoi cells
The randomized points created in step 2 will be
considered as the centers of the Voronoi cells. They
are generated based on the Voronoi tessellation
algorithm inside the box space.
Fig. 3. The overview of the proposed approach 1 to create the random
lattice structure. • Step 4: Creating Voronoi diagrams
• Step 3: Creating Voronoi cells The Voronoi cells include vertex and edges
connecting between vertex. The Voronoi diagrams
The randomized points created in step 2 will be are created based on the vertex and the edges in the
considered as the centers of the Voronoi cells. They cells.
are generated based on the Voronoi tessellation
algorithm inside the box space. • Step 5: Building a frame model
• Step 4: Creating Voronoi diagrams The frame structure is formed by the edges of the
Voronoi diagrams. The part of the frame model
The Voronoi cells include vertex and edges outside the design zone will be trimmed off in order
connecting between vertex. The Voronoi diagrams to fit it into the boundary of the design zone. The
are created based on the vertex and the edges in the frame model of the random lattice structure is
cells. finally designed.
• Step 5: Building a frame model
The frame structure is established by using the
edges of the Voronoi diagrams. In order to fit the
frame inside the boundary of the design space, the
part of the frame structure outside the design zone
is removed. Finally, the frame model including
points and lines is built.
If the density of bars is considered, the second
approach is proposed to modify the distribution of points
in the design zone. The second approach proposes to divide
the design space into different regions by many box spaces.
The density of the lattice structure depends on the number
of points generate randomly in each box space. The global
view of the second approach is shown in Fig. 4. It
comprises the following steps:
Fig. 4. The overview of the proposed approach 1 to create the random
• Step 1: Creating different box spaces lattice structure.
A box surrounding the design space is divided into 3) A 3D geometric model of random lattice structures
different smaller boxes in order to modify the
density distribution of the randomized points in the A 3D solid model of the random lattice structure is able
design space. The number of points in each box to be created by the frame model. Each bar of this lattice
space depends on the density distribution of bars in structure is formed by each line in the frame model. The
the lattice. shape of the bar can be selected as a solid or hollow bar and

978-1-7281-3804-6/19/$31.00 ©2019 IEEE1277 1277

Authorized licensed use limited to: VIT University. Downloaded on February 01,2022 at 15:42:48 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2019 IEEE IEEM

its section could be round, square, rectangular or any shape.

An example of the shape of the bar is shown in Fig. 5.

Solid bar

Round

Square

Rectangular

Fig. 7. The pseudocode algorithm developed in Grasshopper to create a

random lattice structure

Hollow bar

Fig. 5. The geometry of a solid bar with different sections.

III. A CASE STUDY

(a) (b)
In order to applicate the proposed method, an example of Box space Randomized points Voronoi cells
the design and creation of the random lattice structure in
the commercial CAD software as Rhinoceros is presented
in the paper.

A. Design of random lattice structure for helical gear

In this case, the random lattice structure is generated

inside a helical gear in order to illustrate the proposed
method. The 3D geometric model of helical gear and the
design space need to be created as given in Fig. 6. (c) (d)
Voronoi diagrams Frame model

Model of
random lattice
(e)
Design Non-design
Fig. 8. The process to create the random lattice structure.
Fig. 6. A model of helical gear and design space.
A function Populate3D of Grasshopper is used to
In order to generate automatically a geometric model of generate a set of randomized points in a box space as given
the random lattice structure in a CAD environment, a in Fig. 8a. The box space is defined so that it can
Python script function is developed in Grasshopper based completely contain the design space. The function
on the proposed method. A pseudocode algorithm is used Voronoi3D in the library ghpythonlib.components is also
to develop the program in Grasshopper by using Python used in this case to create a volumetric Voronoi diagram
programming language as shown in Fig. 7. based on the randomized points (see Fig. 8b). The diagrams
built in a box space are larger than the design space, so the

978-1-7281-3804-6/19/$31.00 ©2019 IEEE1278 1278

Authorized licensed use limited to: VIT University. Downloaded on February 01,2022 at 15:42:48 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the 2019 IEEE IEEM

part of the diagrams outside of the design zone needs to be ACKNOWLEDGMENT

cut off (see Fig. 8c). A function Intersection in the library
Rhino.Geometry is used to find the intersection between the This work was financially supported by The University
Voronoi diagrams and the design space in order to obtain of Danang Foundation for Science and Technology
the frame structure model of the random lattice structure Development in 2016-2018, research project B2016-
(see Fig. 8d). Each bar of the random lattice structure is DN02-21.
formed by each edge of the frame model using function
AddPipe in the library rhinoscriptsyntax with the value of REFERENCES
the radius. The 3D geometric model of the random lattice
structure is finally built as given in Fig. 8d. [1] R. Tiwari, R. S. Andhare, A. Shooshtari, and M. Ohadi,
“Development of an additive manufacturing-enabled
B. Design of random lattice structure for support compact manifold microchannel heat exchanger,” Applied
structures Thermal Engineering, vol. 147, pp. 781-788, 2019/01/25/,
2019.
[2] M. Zhou, Y. Liu, and Z. Lin, “Topology optimization of
The random lattice structure is able to become an thermal conductive support structures for laser additive
effective and useful tool to create the support structures for manufacturing,” Computer Methods in Applied Mechanics
additive manufacturing thanks to the ability to modify the and Engineering, vol. 353, pp. 24-43, 2019/08/15/, 2019.
density of bars in the lattice structure. An example of [3] S. K. Moon, Y. E. Tan, J. Hwang, and Y.-J. Yoon,
design support structures for a part using additive “Application of 3D printing technology for designing light-
manufacture process is shown in Fig. 9. weight unmanned aerial vehicle wing structures,”
International Journal of Precision Engineering and
Manufacturing-Green Technology, vol. 1, no. 3, pp. 223-
228, 2014/07/01, 2014.
[4] E. Alabort, D. Barba, and R. C. Reed, “Design of metallic
bone by additive manufacturing,” Scripta Materialia, vol.
164, pp. 110-114, 2019/04/15/, 2019.
Surface support Frame model of [5] D. S. Nguyen, and F. Vignat, “A method to generate lattice
structure random lattice structure for Additive Manufacturing,” in 2016 IEEE
International Conference on Industrial Engineering and
Engineering Management (IEEM), 2016, pp. 966-970.
[6] D. S. Nguyen, T. H. T. Tran, D. K. Le, and V. T. Le,
“Creation of Lattice Structures for Additive Manufacturing
in CAD Environment,” in 2018 IEEE International
Conference on Industrial Engineering and Engineering
Management (IEEM), 2018, pp. 396-400.
[7] D. S. Nguyen, “Design of lattice structure for additive
manufacturing in CAD environment,” Journal of Advanced
Mechanical Design, Systems, and Manufacturing, vol. 13,
Density of bars in no. 3, pp. JAMDSM0057-JAMDSM0057, 2019.
support structures [8] M. Fantini, M. Curto, and F. De Crescenzio, “A method to
design biomimetic scaffolds for bone tissue engineering
Fig. 9. The model of the random lattice structure for supports. based on Voronoi lattices,” Virtual and Physical
Prototyping, vol. 11, no. 2, pp. 77-90, 2016/04/02, 2016.
[9] G. Malians, I. T. Sarafis, T. Lazaridis, A. Varoutoglou, and
,V. CONCLUSION G. Tsakataras, “Random lattice structures. Modelling,
manufacture and FEA of their mechanical response,” IOP
The new approaches proposed in the paper permits the Conference Series: Materials Science and Engineering, vol.
product designer to design the random lattice structure and 161, pp. 012045, 2016/11, 2016.
create a 3D solid model of it in the CAD environment. [10] J. Martinez, S. Hornus, H. Song, and S. Lefebvre,
These approaches help to generate the random lattice “Polyhedral voronoi diagrams for additive manufacturing,”
J ACM Trans. Graph, vol. 37, no. 4, pp. 1-15, 2018.
structure with different density distribution of bars in any
[11] F. Aurenhammer, and R. Klein, "Chapter 5 - Voronoi
complex design space based on the Voronoi diagrams. An Diagrams**Partially supported by the Deutsche
example to design and build a 3D model of the random Forschungsgemeinschaft, grant K1 655 2-2," Handbook of
lattice structure for a helical gear is also presented in the Computational Geometry, J. R. Sack and J. Urrutia, eds., pp.
paper. 201-290, Amsterdam: North-Holland, 2000.
In the future, the study will focus on using these [12] M. Tanemura, T. Ogawa, and N. Ogita, “A new algorithm
for three-dimensional voronoi tessellation,” Journal of
lattices as a tool to reinforce the strength of the product and
Computational Physics, vol. 51, no. 2, pp. 191-207,
make the product lighter. Besides, the random lattice 1983/08/01/, 1983.
structure can be used as supports for additive [13] S. Fortune, “A sweepline algorithm for Voronoi diagrams,”
manufacturing processes. Algorithmica, vol. 2, no. 1, pp. 153, 1987/11/01, 1987.

978-1-7281-3804-6/19/$31.00 ©2019 IEEE1279 1279

Authorized licensed use limited to: VIT University. Downloaded on February 01,2022 at 15:42:48 UTC from IEEE Xplore. Restrictions apply.