Today, laser additive manufacturing (LAM) is used in more and more industrial applications. Due to a new freedom in design it offers a high potential for weight saving in lightweight applications, e.g., in the aerospace industry. However, most design engineers are used to design parts for conventional manufacturing methods, such as milling and casting, and often only have limited experience in designing products for additive manufacturing. The absence of comprehensive design guidelines is therefore limiting the further usage and distribution of LAM. In this paper, experimental investigations on the influence of part position and orientation on the dimension accuracy and surface quality are presented. Typical basic shapes used in lightweight design have been identified and built in LAM. Thin walls, bars, and bore holes with varying diameters were built in different orientations to determine the process limits. From the results of the experiments, comprehensive design guidelines for lightweight structures were derived in a catalog according to DIN 2222 and are presented in detail. For each structure a favorable and an unfavorable example is shown, the underlying process restrictions are mentioned and further recommendations are given.

Laser additive manufacturing (LAM) is a repeating layer wise manufacturing process which uses a laser beam to melt and solidify material in a powder-bed according to slices of a corresponding three-dimensional computer aided design (3D-CAD) model. Stepwise production causes a reduction of complex 3D geometries into simpler two-dimensional manufacturing steps.1–5 

Lightweight design studies, e.g., in the field of aviation, already successfully demonstrated LAM's high potential for this field of application due to its geometrical freedom. Weight reductions of up to 80% compared to conventionally manufactured counterparts were feasible.6–13 However, these studies were generally created by experienced designers as the process inherent restrictions are not yet widely spread. Even though LAM's geometrical freedom can be significantly higher compared to conventional manufacturing approaches, it has process specific restrictions. These strongly differ from conventional processes and must be considered when designing parts.

Creating design guidelines are thus of uttermost importance in order to allow a wide industrial usage of LAM. Available studies on design guidelines for LAM currently only give recommendations regarding the basic process inherent restrictions, influences, part design as well as general post processing.14–19 A comprehensive overview is not yet available. Comparable approaches have been published by Adam and Zimmer20 with a focus on general guidelines for metal (steel and aluminum) and synthetic material based processes, while this paper focusses specifically on lightweight structures made of TiAl6V4.

In order to narrow the field of initial research, lightweight applications in TiAl6V4 were in the focus of this first guideline generation approach due to the high potential of LAM in this field of application.10–13 A robust part design and less the processes full exploitation was the goal. To identify the most basic shapes often occurring in such structures, common guidelines for general lightweight design were used (cf. Klein).21,22 Typical manufacturing challenges in lightweight design include: general part accuracy, bores, surface quality, beam elements, final machining, hollow structures, walls/rib structures, and LAM specific support structures. Dimensions of the test specimens follow the lightweight paradigm of a minimal design. Focussing on a minimal design approach moreover facilitates a reduction in manufacturing time and costs.

Test specimens used for guideline generation were consistently manufactured on unmodified industrial production machines allowing transferability for a broad field of users. The specimens design target was to ease measuring and later analysis. Holistic test parts incorporating multiple test specimens were thus avoided. Manufacturing was conducted on an EOS 270xt and Concept M2 using a carbon brush recoater device and the alloy TiAl6V4, see Table I.

TABLE I.

Basic process parameters of LAM machines used.

Machine EOS M270 XT Concept laser M2
Material  TiAl6V4, retail powder  TiAl6V4, retail powder 
Layer thickness (μm)  30  30 
Laser power (W)  Core: 170; skin: 120  Core: 180; skin: 200 
Scan velocity (mm/s)  Core: 1250; skin: 1000  1250 
Hatch distance (mm)  0.1  0.15 
Exposure type  Stripe  Stripe 
Machine EOS M270 XT Concept laser M2
Material  TiAl6V4, retail powder  TiAl6V4, retail powder 
Layer thickness (μm)  30  30 
Laser power (W)  Core: 170; skin: 120  Core: 180; skin: 200 
Scan velocity (mm/s)  Core: 1250; skin: 1000  1250 
Hatch distance (mm)  0.1  0.15 
Exposure type  Stripe  Stripe 

First tests included part accuracy analysis in dependence of building plate positioning, recoater related orientation and part size itself. The intention was to identify possible influences to be considered for later test specimen setup. Adjacent to this further test specimens focusing on surface quality and basic geometries in lightweight design were analyzed. Basic geometries examined include thin walls, bar elements as well as bore holes.

Results were incorporated into a design catalogue. Design catalogs allow an increase of solution variety and an ease in its synthesis at frequently occurring activities in product development and design.23 Especially in fields of recurring design challenges such as manufacturing restrictions, the creation of design catalogues can support an efficient product development. The catalog presented is created in dependence on VDI 2222.24 For clearness, it is separated in several columns. The classification section on the left is divided according to basic geometries, part design, and final machining. The sections are colored in order to ease handling. The main part incorporates pictures showing preferable design solutions according to the manufacturing challenge. Columns on the right side show remarks and more detailed information on the specific restriction.

This section summarizes the most basic experiments conducted in order to set up design guidelines for LAM. The examinations focus was TiAl6V4 and lightweight structures, see Sec. II.

Especially for end-use parts, accuracy is often vital for fulfilling functionalities. In lightweight design for LAM there is a striving toward thin walled and highly complex structures in order to save weight and reduce manufacturing time. The smaller these structures become, the more severe are the consequences in an accuracy deviation of structural relevant parts as the mechanically loadable cross sections deviate from the intended design. A part failure below the intended loading scenario can be the result. Analyzing the process influences regarding the accuracy in LAM can help the designer to already consider this behavior in the early design stage. Experiments include the analysis of the influence of part position as well as size and orientation on the accuracy of basic test geometries.

1. Position dependency of LAM accuracy

The first influence on the part accuracy that is evaluated is the parts positioning on the building platform. Influences leading to deviations can be the mechanical and optical machine setup or the powder quality, just to name a few. Five part positions were analyzed: M = middle, UR = upper right, UL = upper left, LR = lower right, and LL = lower left. The designed test specimen, see Fig. 1 incorporates three basic cross sections with a building height of 10 mm and a connection base plate that incorporates the location marker. The dimensions of the specimens are modeled in such a way that a robust manufacturing is guaranteed. The cylinder has a diameter of 5 mm, the rectangular a homogeneous side length of 5 mm, and the wall a length of 10 mm and a thickness of 0.5 mm. All test specimens were heat treated in order to prevent inaccuracies to thermally induced stresses in the base plate and wire eroded to ease measuring.

FIG. 1.

Test geometry for the position dependency of the accuracy: shape and positioning; measuring directions indicated blue; planes of measuring indicated orange.

FIG. 1.

Test geometry for the position dependency of the accuracy: shape and positioning; measuring directions indicated blue; planes of measuring indicated orange.

Close modal

Measuring of the test specimens was conducted in three ways: (1) vernier caliper; (2) microscope, probes as build; and (3) microscope, specimens embedded and polished. The intention was to identify the most suitable way of measuring test parts in terms of time and cost efforts. Table II shows an excerpt of the measuring campaign. According to Fig. 1, two perpendicularly measure points were taken per sample at three different heights.

TABLE II.

Position dependency test specimen: excerpt of measuring results.

      Vernier caliper measuring results (mm)
      Position on building plate
  Measuring position Measuring direction UL UR M LL LR
Cylinder (Ø 5 mm)  Top  5.01  5.05  5.05  5.11  5.07 
   5.06  5.08  5.05  5.07  5.04 
Middle  5.08  5.01  5.02  5.05  5.07 
   5.05  5.05  5.02  5.05  5.03 
Bottom  5.04  5.06  5.03  5.06  5.04 
   5.02  5.05  5.04  5.08  5.02 
      Microscope measuring results (mm) 
Top  5.09  5.08  5.04  5.07  5.06 
   5.03  5.09  5.06  5.07  5.06 
      Microscope measuring results; specimen embedded and polished (mm) 
Top  5.10  5.08  5.06  5.10  5.15 
   5.11  5.14  5.08  5.10  5.13 
      Vernier caliper measuring results (mm)
      Position on building plate
  Measuring position Measuring direction UL UR M LL LR
Cylinder (Ø 5 mm)  Top  5.01  5.05  5.05  5.11  5.07 
   5.06  5.08  5.05  5.07  5.04 
Middle  5.08  5.01  5.02  5.05  5.07 
   5.05  5.05  5.02  5.05  5.03 
Bottom  5.04  5.06  5.03  5.06  5.04 
   5.02  5.05  5.04  5.08  5.02 
      Microscope measuring results (mm) 
Top  5.09  5.08  5.04  5.07  5.06 
   5.03  5.09  5.06  5.07  5.06 
      Microscope measuring results; specimen embedded and polished (mm) 
Top  5.10  5.08  5.06  5.10  5.15 
   5.11  5.14  5.08  5.10  5.13 

The absolute deviations are in the field of the expected surface roughness according to the machine manufacturers specifications.25 All parts dimensions were above the designed dimensions, see Table II. The results show that the part position has no significant influence on the part accuracy on the basis of the chosen machines. Even though the measured dimensions differ from the designed CAD-model up to +0.15 mm, no influence of the part position on this deviation can be determined. Other machine setups, especially in terms of optical concepts, may lead to different results due to different focal characteristics. Moreover the selected measuring procedures show comparable results as well. The vernier caliper will be the measuring device of choice for the following campaigns where suitable due to the ease of measuring.

2. Size and recoater related orientation dependency of LAM accuracy

LAM uses metal powder as raw stock. Manufacturing a part according to the CAD data thus makes the layer wise melting and solidification of the melt pool necessary. Possible part shrinkage, especially depending on the part dimensions, can be the result. Thus test specimens were designed in order to assess the dependency of the accuracy on part size and machine capabilities regarding shrinkage compensation. The testing included rectangular, cylindrical, and elliptical cross sections of a 35 mm build height. Dimensions of rectangular test specimens include 0.5, 1, 2, 5, 10, and 15 mm width parts at lengths of 30 and 60 mm. Cylindrical specimens incorporate outer diameters of 3, 6, 15, 24, and 30 mm at wall thicknesses of 2 mm. Tested ellipses incorporate major axis of 15, 20, and 30 mm, a minor axis of 7.5, 10, and 15 mm as well as a wall thickness of 2 mm. Manufacturing was conducted on a Concept Laser M2.

Parts were built in three angular alignments toward the recoater device at a height of 30 mm: 0°, 45°, and 90°. Manufacturing was conducted on a Concept Laser M2, see Sec. II. Focus of measurements was the part length and width as built on the thermally not treated baseplate by means of a tactile coordinate measuring system.

Measurements were taken at three different heights of the part, see Fig. 2. The thickness was measured on 10 points along the longitudinal axis of the part for the 30 mm as well as the 60 mm long rectangular specimens. The circular and elliptical parts were analyzed in terms of shape accuracy. In previous studies cylindrical thin walled cross sections tended to bend inside, leading to an insufficient dimensional accuracy.26 

FIG. 2.

Average deviation from designed width of rectangular test specimen depending on height location of measurement.

FIG. 2.

Average deviation from designed width of rectangular test specimen depending on height location of measurement.

Close modal

The results of the rectangular specimen's wall thickness and length all lied in a range of 0–0.1 mm above the designed geometry, see Fig. 2. The cylindrical cross sections width showed the same trend and no sign of bending in or outward. The test specimens showed that the accuracy of simple robustly designed parts is repeatable and not depending on the parts size if built perpendicular to the building platform according to the test specimens size range.

Compared to conventional manufacturing, the orientation dependency of the achievable surface roughness in LAM is a process specific feature that needs to be considered when designing parts. In general, the higher the surfaces angular alignment toward the building platform gets, the higher the roughness, see Fig. 3. Surpassing a critical angle leads to an increased droplet effect on down facing surfaces and finally to an erroneous buildup of the layers due to a reduced heat flux into the powder bed compared to completely solidified metal. An abortion of the process and in general a loss of the part is the effect. The use of support structures can counteract this behavior and is recommended above the highly material and manufacturing machine depending critical angle. Currently, various studies are available including materials like steel alloys or aluminum.3,14–19,27 In terms of TiAl6V4 no investigations on state of the art machines are available. Manufacturing was conducted on an EOS M270XT.

FIG. 3.

Surface roughness test specimens; top: cube for upfacing surfaces; bottom: bridge for downfacing surfaces (a = plane of process failure).

FIG. 3.

Surface roughness test specimens; top: cube for upfacing surfaces; bottom: bridge for downfacing surfaces (a = plane of process failure).

Close modal

Built up test specimens are shown in Fig. 3 and incorporate sets for upfacing and downfacing surfaces. Surface quality is highly depending on powder grain size distribution, layer thickness and energy input in general (cf. Ref. 28). Measurement was conducted by a perthometer. Even though absolute values will differ with changing machine setups, the tests goal is to determine general trends and compare them with the manufacturer's specifications.

Figure 4 shows the surface roughness of up facing built test specimens in dependence of the angular alignment. Measurement was conducted parallel and orthogonal toward layer plane showing comparable results. Staircase effect and exposure patterns lead to slightly higher roughnesses in orthogonal orientations. At 0° orientations surface roughness shows the lowest values with Rz = approx. 20 μm, see Fig. 6. Roughness significantly rises with the angular alignment showing comparable values between differing orientations of Rz = approx. 60–70 μm. Variations in results can occur due to powder particle adhesion or variations in local powder grain size distributions.28 

FIG. 4.

Roughness of up facing surfaces vs angular alignment toward building platform; orthogonal/parallel indicate measurement orientation toward exposure direction.

FIG. 4.

Roughness of up facing surfaces vs angular alignment toward building platform; orthogonal/parallel indicate measurement orientation toward exposure direction.

Close modal

Figure 5 shows the down facing areas surface roughness in dependence of the angular alignments. Angular alignment is measured toward the building platforms normal vector (cf. Fig. 4). A 90° orientation leads to a process failure and was not manufacturable. Surface roughness shows an explicit increase at angular alignments above β = 50°. With Rz = approx. 50–70 μm from 10° to 50°, surface roughness is increasing from β = 50° to 80° up to approx. 130 μm. Layer wise build up leads to larger overhanging layer sections with rising angular alignments resulting in an increase in powder adhesion due to a meltpool larger than one layer thickness.29 

FIG. 5.

Roughness of down facing surfaces vs angular alignment toward building platform.

FIG. 5.

Roughness of down facing surfaces vs angular alignment toward building platform.

Close modal

In terms of buildability, support structures are necessary at angular alignments above β = 80°. Yet, surface roughness significantly rises above angular alignments of β = 50°, see Fig. 5, which must be considered when aligning parts. Overhanging part sections shall be avoided. In general up facing surfaces show better surface qualities than down facing surfaces (cf. Figs. 4 and 5).

1. Thin walls

Lightweight structures often incorporate thin walled sections, e.g., flanges or shear areas. Available studies and recommendations for manufacturable wall thicknesses only refer to structures perpendicular to the building platform. However, the design of functional parts requires information for manifold orientations in order to exploit the freedom of design and guarantee a robust design.

Figure 6 shows the test specimens on the building platform as built. They are the basis of the analysis and were not subject to any postprocessing. The specimens' angular variation to the recoater device, see Fig. 6, was defined at angles of α = 0°, 45°, 90°, 135°, and 180°. Its angular variation versus the building platform was defined at angels of β = 30°, 45°, 60°, and 90°. Data provided by EOS25 specifies the minimum wall thickness from 0.3° to 0.4 mm. Detail on the structures orientation is not stated. Due to the fact that the orientations influence on the walls manufacturability was difficult to predict, a variation of wall thicknesses was conducted. Built thicknesses were 0.2, 0.3, 0.4, 0.5, 0.6, and 0.7 mm.

FIG. 6.

Orientation of test specimens: α: angle vs the recoater device; β is the angle to the building platform.

FIG. 6.

Orientation of test specimens: α: angle vs the recoater device; β is the angle to the building platform.

Close modal

Specimens with thickness of 0.2 mm did not show a stable process behavior. The maximum height varied from 3 to 5 mm. Of the specimens with a thickness of 0.3 mm only the ones with an orientation to the building platform of β = 30° were fully built.

Incompletely manufactured walls with a thickness of 0.3 mm all incorporated a double walled structure (cf. Ref. 26). At an average height of 6 mm two walls at a thickness of 0.3 mm emerge out of the original wall. All walls exhibiting this effect were not built to full height. Moreover, some of the specimens with thickness of 0.2 and 0.3 mm were bent opposite the originally defined orientation to the building platform, see upper left area of the building platform on Fig. 6. The effects origin is a rise in thermally induced stresses. Walls with thickness of 0.4 mm and above were successfully built.

Experimental results show that it is possible to manufacture thin walled structures made of TiAl6V4 at all examined variations of orientation, see Fig. 6, at minimum wall thickness from 0.4 mm upward.

Figure 7 shows the influence of the orientation to the recoater device and the building platform on the overall achieved deviations. According to Fig. 7 an increase of the wall thickness with a decreasing angle toward the building platform can be observed. In general the highest deviation can be observed at an angle of 30° toward the building platform. Moreover an orientation against the motion direction of the recoater device can lead to an additional reduction of the deviation.

FIG. 7.

Wall thicknesses deviation depending on the angular alignment vs the recoater device (α) and the building platform (β).

FIG. 7.

Wall thicknesses deviation depending on the angular alignment vs the recoater device (α) and the building platform (β).

Close modal

The reason for the increase of the wall thickness with decreasing angle toward the building platform can be found in the processes of layer wise manufacturing leading to an increase in the deviation between the measured and the defined thickness, see Fig. 7. This effect is also known as staircase effect.

Furthermore one reason for successful buildup of 0.3 mm thick walls at an orientation of 30° can be found in this effect. Increase in thickness leads to a gain in structural stability as well. The exposed single layers cover an overall larger area. Thus a better connection to lower and already exposed layers than at the 0.3 mm thick walls with different orientations is the result. Furthermore, the moment acting on the wall and induced by the recoater device exhibits its lowest value at an angle of 30° to the building platform due to the lowest overall height perpendicular to it.

Moreover, the staircase effect is connected with a partial overhang of the layer to be exposed leading to an increase of direct heat flux in the powder bed and thus an increased powder adhesion.1,2,29

2. Bar elements

In terms of lightweight design framework structures consisting of bar elements are very efficient. In addition, these structures can be used for strengthening conventional support structures at sections incorporating high residual stresses or high build heights. Observed test specimens incorporate bars of 0.5, 1, 1.5, and 2 mm diameter at a length of 10 mm and orientations adverse the building platform of 30°, 45°, 60°, and 90°. The angular alignment was consistently along the recoater direction in order to minimize recoating forces on the structure. In order to extend the manufacturing limits, further additional pins of a thickness of 0.5 mm at a length of 80 mm and an orientation of 30°, 45°, 60°, and 90° were built. Test specimens were not subjected to any post processing. Manufacturing was conducted on an EOS M270XT. The pin specimens were fully built without visual signs of defects or deviation showing that beam elements with a thickness of 0.5 mm and above can be successfully built according to the test parameters.

3. Bore holes

Bores are an essential design feature. Engineering parts, especially lightweight structures addressed by the guidelines presented, often incorporate interfaces in form of rivets or screws requiring bores in the parts to be connected, see also Sec. II. Recent developments further showed the potential for an integration of functions in form of internal cooling channels or fluid ducting.9 In order to generate guidelines for the outlined structures, two test specimens were built, see Fig. 8 for vertical bore orientation. The parts high volume made a stress releasing thermal treatment necessary in order to avoid size deviations by structural distortion. Manufacturing was conducted on an EOS M270XT.

FIG. 8.

Test specimen for bore diameters; build direction perpendicular toward building platform.

FIG. 8.

Test specimen for bore diameters; build direction perpendicular toward building platform.

Close modal

Dimensions of the bore test specimens are 59 × 20 × 28.5 mm, see Fig. 8. Incorporated bore diameters are shown in Fig. 9. Specimens were built with orthogonal and parallel orientation of the bore axes toward the building plate considering extreme positioning. Measured bore diameters all were throughout larger than designed, see Fig. 8.

FIG. 9.

Parallel and orthogonal oriented bores: deviation from designed dimensions in dependence of bore diameter.

FIG. 9.

Parallel and orthogonal oriented bores: deviation from designed dimensions in dependence of bore diameter.

Close modal

Bore diameters of 2 and 3 mm incorporated adhering powder after manufacturing that could only be removed by force. Bores above 3 mm did not incorporate this effect. Low bore dimensions may lead to lower temperature flux compared to larger ones leading to the described adhesion and bore blocking effect. The highest accuracy can be achieved with orthogonal built bore geometry features, see Fig. 9. All bores of the parallel built specimen show a material fall in on the bores upper side due to a curling of the upper sections of the bore shortly before connecting the layers that close the bores cross section (cf. Ref. 19). Strategies for avoiding these effects by designing bores that outrun sharply to the top were developed by Thomas.19 Moreover results can be used for the design of larger hollow structures as well. Manufacturing of process stable hollow structures is often limited by the effective radius at the top leading to insufficient manufacturing characteristics of horizontally oriented circular cross sections. Already successfully implemented in lightweight design is an alternative elliptical cross section with an effective upper structural radius below process critical dimensions allowing larger moments of inertia, see Table III.

TABLE III.

Guideline catalogue for laser additive manufacturing of TiAl6V4.

graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 
graphic
 

Widely spread is a conception of LAM as a post machining free manufacturing process. Yet building TiAl6V4 parts on building platforms requires later separation. High demands on surface roughness and part tolerances can additionally oblige the designer to consider conventional manufacturing processes for post machining. Their process inherent restrictions can significantly influence the design freedom. Incorporating the most basic design guidelines for conventional manufacturing processes into a design guideline for LAM thus is a necessity as it raises the designer's awareness of the holistic manufacturing chain. Typical post machining processes for lightweight brackets primarily incorporate wire eroding and milling.

Laser Additive Manufacturing offers a unique freedom in design, yet differs greatly from conventional manufacturing in terms of operating principles. Although theoretical benefits in design are apparent, little research in design guidelines has been conducted yet. Multiple design iterations until a robust building process and an only marginally exploitation of the design possibilities are the result.

Thus basic experiments exploiting manufacturing restrictions for basic geometries common in lightweight design are being presented in this paper. A structured catalogue for basic design guidelines focused on LAM of TiAl6V4 is delivered, offering designers the possibility to consider LAM restrictions from the early stage of design.

In order to fully exploit the geometrical freedom of the process and broaden its industrial application it is of uttermost importance to continue the development of design guidelines for LAM. Therefore further process specific restrictions will be explored and summarized in a simple and intuitive way offering designers and inexperienced users of LAM a comprehensive overview of its limitations and possibilities.

The research presented was mainly conducted in the research project TiLight which was funded by the “Bundesministerium für Bildung und Forschung” under the support code 03CL20A. Moreover, partial results were researched during the IGF Research project AlaTin (00.41 Z) of the Forschungsvereinigung DVS—Deutscher Verband für Schweißen und verwandte Verfahren e.V. It was funded by the AiF under the “Bundesministerium für Wirtschaft und Energie.”

1.
A.
Gebhardt
,
Generative Fertigungsverfahren
(
Carl Hanser Verlag
,
München
,
2007
).
2.
M. F.
Zäh
,
Wirtschaftliche Fertigung mit Rapid-Technologien
(
Carl Hanser Verlag
,
München
,
2006
).
3.
W.
Meiners
, “
Direktes Selektives Laser Sintern einkomponentiger metallischer Werkstoffe
,” Ph.D. thesis,
RWTH Aachen
,
1999
.
4.
C.
Over
, “
Generative Fertigung von Bauteilen aus Werkzeugstahl X38CrMoV5-1 und Titan TiAl6V4 mit selective laser melting
,” Ph.D. thesis,
RWTH Aachen
,
2003
.
5.
C.
Emmelmann
,
J.
Kranz
,
D.
Herzog
, and
E.
Wycisk
, “
Laser additive manufacturing of metals
,” in
Laser Technology in Biomimetics
(
Springer
,
Berin, Heidelberg
,
2013
).
6.
SAE International,
Aerospace Engineering & Manufacturing
(
SAE International
,
Warrendale
,
2010
), Vol.
2
, No. 29.
7.
J.
DeGrange
, “
Boeing's vision for rapid progress between dream and reality
,” in
Euro-uRapid
, Frankfurt, 27–28 November, 2006.
8.
J.
DeGrange
, “
Steps to improve direct manufacturing readiness levels
,” in
Euromold 2006
, Frankfurt, 29 November–2 December (2006).
9.
T.
Wohlers
,
Wohlers Report 2013—State of the Industry
(
Wohlers Associates
,
Fort Collins, CO
,
2013
).
10.
R.
Hague
,
I.
Campbell
,
P.
Dickens
, and
P.
Reeves
, “
Integration of solid freeform fabrication in design
,” in
Proceedings of the Solid Freeform Fabrication
, University of Texas, Austin, 6–8 August, 2001.
11.
D. M.
Watts
and
R. J.
Hague
, “
Exploiting the design freedom of RM
,” in
Proceedings of the Solid Freeform Fabrication
, University of Texas, Austin, 14–16 August, 2006.
12.
C.
Emmelmann
,
P.
Sander
,
J.
Kranz
, and
E.
Wycisk
, “
Laser additive manufacturing and bionics: Redefining lightweight design
,”
Phys. Proc.
12
,
364
368
(
2011
).
13.
C.
Emmelmann
,
M.
Petersen
,
J.
Kranz
, and
E.
Wycisk
, “
Bionic lightweight design by laser additive manufacturing (LAM) for aircraft industry
,”
Proc. SPIE
8065
,
80650L
(
2011
).
14.
L.
Castillo
,
Study About the Rapid Manufacturing of Complex Parts of Stainless Steel and Titanium
(
TNO Industrial Technology
,
Delft
,
2005
).
15.
J. P.
Kruth
,
B.
Vandenbroucke
, and
P.
Van Vaerenbergh
, “
Benchmarking of different SLS/SLM processes as rapid manufacturing technologies
,” in
International Conference of Polymers and Moulds Innovations (PMI) Gent
,
2005
.
16.
J. P.
Kruth
and
B.
Vandenbroucke
, “
Selective laser melting of biocompatible metals for rapid manufacturing of medical parts
,”
Rapid Prototyping J.
13
,
196
203
(
2007
).
17.
C.
Aumund-Kopp
and
F.
Petzold
, “
Laser sintering of parts with complex internal Structures
,” in
PM World Congress, Fraunhofer Institute for Manufacturing Technology and Applied Materials Research
, Bremen,
2008
.
18.
O.
Kushnarenko
,
Entscheidungsmethodik zur Anwendung generativer Verfahren für die Herstellung metallischer Endprodukte
(
Shaker Verlag
,
Aachen
,
2009
).
19.
D.
Thomas
, “
The development of design rules for selective laser melting
,” Ph.D. thesis,
University of Wales, Cardiff
,
2009
.
20.
G.
Adam
and
D.
Zimmer
, “
Design rules for additive manufacturing—element transitions & aggregated structures
,”
CIRP J. Manuf. Sci. Technol.
7
(
1
),
20
28
(
2014
).
21.
B.
Klein
,
Leichtbau-Konstruktion: Berechnungsgrundlagen und Gestaltung
(
Vieweg + Teubner
,
Wiesbaden
,
2009
).
22.
J.
Wiedemann
,
Leichtbau—Elemente und Konstruktion
(
Springer-Verlag
,
Berlin, Heidelberg
,
2007
).
23.
K.-J.
Conrad
,
Taschenbuch der Konstruktionstechnik
(
Carls Hanser Verlag
,
München
,
2004
).
24.
Verein
Deutscher Ingenieure
,
VDI Richtlinie 2222: Konstruktionsmethodik—Erstellung und Anwendung von Konstruktionskatalogen
(
Verein deutscher Ingenieure, Beuth Verlag
,
Berlin
,
1982
).
25.
EOS GmbH,
Technisches Datenblatt–Laser-Sinter-System EOSINT M 270
(
Electro Optical Systems
,
Krailling/Münchenk
,
2009
).
26.
J.
Kranz
, “
Manufacturing restrictions for laser additive manufacturing of lightweight structures made of TiAl6V4: Thin wall structures
,” in
Proceedings of DDMC Direct Digital Manufacturing Conference 2012
, Berlin, 14–15 March, 2012.
27.
G.
Strano
, “
Surface roughness analysis, modelling and prediction in selective laser melting
,”
J. Mater. Process. Technol.
213
,
589
597
(
2013
).
28.
V.
Seyda
, “
Investigation of aging processes of Ti-6Al-4V powder material in laser melting
,”
Phys. Proc.
39
,
425
431
(
2012
).
29.
D.
Wand
, “
Study on the designing rules and processability of porous structure based on selective laser melting (SLM)
,”
J. Mater. Process. Technol.
213
,
1734
1742
(
2013
).