Investigation On Mechanical Properties of Wire Arc Additive Manufacturing (WAAM) of Stainless Steel Using FEM Simulation and Experiment

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 18

Investigation on Mechanical

Properties of Wire Arc


Additive Manufacturing
(WAAM) Of Stainless Steel
Using FEM Simulation and
Experiment
Student: Mohamad Aiman Bin Yazid
UITM ID: 2015475449
Supervisor: Prof. Dr. Ing. Yupiter Hp Manurung

ABSTRACT
Wire Arc Additive Manufacturing (WAAM) process forms a part with
complex geometry over the layer-by-layer deposition of welded beads, it is
important to model the geometry of a single weld bead as well as the multi-
bead overlap process to achieve high surface quality and dimensional
accuracy of manufactured parts. This study produces 316L stainless steel
parts through experimental set-up and FEM simulation. Simulation will
focus on the comparison of all temperature strain hardening model to high
temperature for stainless steel 316L material. The experimental specimen
will be term of welding quality on the profile, material composition and
simulated by using FEM simulation. On simulation, the best parameters that
perform as the manipulated variables that are welding speed, the current
used for welding, the wire feed rate and the cooling time required for each
layer that are responsible for the quality and changes in the microstructure
characteristics of the WAAM products. The specimen was prepared to find
the quality of the welding by Vickers Hardness test. Macrostructure also
required to calibrate size of every weld filler between simulation and
experimental. The chemical composition of the wire and the specimen was
identifying for ensure both are match by using the Energy Dispersive X-Ray
Analysis (EDX) machine. The residual stress of the WAAM also discover by
using X-Ray Diffraction (XRD) for comparison experimental and FEM
simulation using MSC Marc Mentat software. All values were analyzing and
discussed the new development of 316L Stainless steel on Wire Arc Additive
Manufacturing (WAAM).
1.0 Introduction
Wire arc addictive manufacturing (WAAM) was combination of an electric
arc heat source and wire as feedstock. The first pattern of wire arc additive
manufacturing (WAAM) was filed in 1925, however only since 1990s it has
been investigated for additive manufacturing (AM) purpose [1]. Additive
Manufacturing (AM) allows for increased design freedom, potentially
leading to weight savings as well as facilitating the manufacture of complex
assemblies that used to consist of many subcomponents. As a rapid
prototyping technology, Additive Manufacturing (AM) is a capable
manufacturing ways building objects layer by layer[2]. Equipped with
traditional manufacturing methods, the AM method can cost-effectively save
processing time, reduce post-processing time, also reduce material waste and
enhance material utilization. Wire and arc additive manufacturing (WAAM)
adapts the multi-layered welding as its basic operating model, similar to a 3D
printing technique. [5]
In this study, AISI 316L has the best weld ability among stainless
steels and low carbon content, carbide precipitation has less chance to occur
during welding. The mechanical properties testing analysis will show the
strength and elongation result at the forged level and the parts can meet the
application requirements. Then there are also the mechanical properties of
anisotropy, which qualities differ in different ways from the result of the
consolidation of the grain boundary.
Finite element process simulation arranges for an efficient way to
study modification strategies for certain problem. A WAAM modelling
strategy is anticipated created on an original heat source model that takes into
interpretation the actual power supply between filler and base materials [3].
In order to verify the effectiveness of proposed modelling, an experimental
validation is delivered by comparing the measured distortions of a WAAM
tests-case with the simulated ones, emphasizing the accuracy of proposed
model. In our case, regarding the performance this project, how a system of
robotic welding can help in welding and define the welding parameters for
the specimen. Instead, the Wire Arc Additive Manufacturing (WAAM)
process using experimental setup and FEM simulation involving the welding
quality (e.g. Surface finish, Chemical composition, Hardness Test, Effective
stress and Macrostructure characteristics). The comparison of the simulation
and experimental setup will focus on all temperature strain hardening model
to high temperature for stainless steel material.
2.0 METHODOLOGY
On this project, we compulsion to conduct the experimental to acquire the
parameter before proceed on FEM simulation of wire arc additive
manufacturing (WAAM).
2.1 Mechanical behavior of Wire AISI 316L stainless steel

In our experiment, 316L Stainless Steel was the purpose of the material
study. Austenitic stainless steels, such as AISI type 316, are especially
interesting for WAAM because they are relatively expensive to machine.
Near net shaped structures could be manufactured without appreciable
material losses, decreasing the waste of expensive metals such as Ni, Cr and
Mo. For engineering applications, AISI 316 steel presents outstanding
intergranular corrosion resistance, good grain corrosion resistance to most
chemicals, salts and acids and Mo content helps increase resistance [4]. The
low carbon version of AISI 316, called 316L, is virtually immune to
sensitization (grain boundary carbide precipitation).
2.2 Experimental using Robotic Welding
Austenitic Cr-Ni Stainless Steel 316L was the material to be used. Before
finding the optimum welding parameters, many specimens need to be used.
The size of the specimen was prepared using a mild steel type of 100 mm x
200 mm x 9 mm. This process will be conduct by ABB GMAW Robotic
Welding located at the Robotic Welding Workshop. The path planning is an
important step. Parameters of the process such as:
A. Wire feed rate
B. Path of deposition
C. Speed of deposition
D. Power supply

In mechanical engineering, the following mathematical formula is


used to calculate how much heat is required for welding and each calculation
allows users to know how to manually perform such calculations.
Mathematical Formula P = Power of heat input
of Welding heat input: - V = Voltage (V)
I = Current (A)
P = V x I x 0.06/s S = speed (s)
All parameter has to be carefully planned in order to achieve the best
possible geometrical accuracy, surface quality, and mechanical performance
of the fabricated parts.
2.3 Drawing of Simulation

Open source software MSC Marc/Mentat will be used to draw the


simulation. The simulation results will be used to make comparison with
experimental welding process. The steps required to draw the simulation;
1. Geometric modelling and FEM meshing
2. Assigning clamping and bearing
3. Assigning materials
4. Determining the weld path line
5. Defining welding parameters
6. Heat source modelling
7. Load case setting and set the tracking points
8. Running the simulation
9. Result and post-processing

Since the WAAM simulation requires a multi-pass welding


sequence, the numbers of similar procedures as above should be repeated
depends on how much the layer and string added in FEM WAAM
simulation. In this research, 3 String and 10 layers weld beads is
implemented, which indicates that numbers of procedure that should be
repeated is 10 times before move on to the load case setting.

Figure 2.3.1 : Workspace for WAAM simulation at MSC Marc Software

On this simulation, the finishing product must calibrate by several


comparisons. It was identifying the several result which are: -
a) Distortion
b) Effective stress (residual stress) on the normal and high temperature
c) Equivalent von misses
d) Temperature increment
2.4 Inspection and Testing of 316L Stainless Steel in WAAM
The experimental result required cutting form a surface toward
undergo metallographic evaluation hardness test and micro-structure
visualize. One cross section of the wall from on weld filler specimen was cut
perpendicular to the layer length direction for inspection, Vickers pyramid
hardness measurements (HV, with 1 kg force) and macrostructure
characterization. These samples were subjected to standard metallographic
techniques such as grinding and polishing, as well as etching for point
counting of the size weld filler in the first and the last bead. In order to obtain
the mechanical properties of the parts, a hardness test was conducted to
investigate the strength of the parts.
After that, specimen was accompanied through macrostructre
machine for define the size of every multibead weldfiller for comparison on
the FEM simulation. Each weldbead size can be distinguished easily by the
etching process. The macrostructure identifies and adjusts some focus by
large scale. Manual method of measurement of the size weld filler was used
by inspecting the visual size by the standard ruler for measurement.

Figure 2.4.1 : Macrostructure machine


Used vicker harness testing machine implemented the hardness test.
The hardness test to determine by measuring the penetration depth of the
indenter or by measuring the size of an indenter's left impression. The
Vickers hardness testing capable applied 1 kg load to every weld bead. The
test result must be able to compare the physical hardness test measurement.

Figure 2.4.2 : Vicker Harness Figure 2.4.3 : Calculation


Test Machine of Vicker hardness test
2.5 Chemical Composition Wire 316L stainless steel and
Specimen of WAAM
Following welding completion, the chemical composition of the
specimen shall be identical to the wire feed on the 316L stainless. As we
know, the process of welding the WAAM and stainless steel wire 316L has
to be performed in two specimens. The entire specimen has been cut by used
water jet machines and a thickness measurement of about 25 mm for the
welding process specimen. But the actual 316L wire was hard to detect on
these machines physically. Therefore, the wire was mounted by using
Simplimet 2000. In order to make the clear surface available for the
determination of the wire, a certain process is needed that is grinding,
polishing and etching
Set the two specimens machined and carried out with the support of
a laboratory technician on the EDX machine. This machines stimulated and
we had to choose six points to proceed randomly with the specimen. This
requires us to identify the chemical composition to be identified in the
experiment.

Figure 2.5.1 : Energy-dispersive X-ray


spectroscopy(EDX) Machine

2.6 Determine The Residual Stress On the Specimen WAAM

Residual stress is the distribution of internal stress that is locked into a


material. These stresses occur even after all external loading forces have
been removed. The residual stress with references should be determined by
the WAAM specimen process. On X-Ray Diffraction machines (XRD), the
sample must be setup. In this situation, the process of roughly cutting must
be applied to the machines to be fixed. Non-traditional machines locate the
cutting toll, then finalize using water jet because the less stress may occur on
the specimens. The residual stress on compressive and tensile stress was
usually left by the process welding and machinist. The nine point was placed
in the specimen on every 3 layers of each string to find the residual stress.
3.0 RESULT AND DISCUSSION
3.1 WELDING PROCESS
ABB GMAW Robotic Welding located at the Robotic Welding Workshop
will perform this process. It is important that the welding parameters such as
current (A), wire feed (m / min), voltage (V) and welding speed (mm / s) are
carefully selected. The WAAM experimental with SS316L started with
finding the best welding parameters as well as finding out what geometry the
weld seam has with the various welding parameter. The welding parameters
used are shown in the table: -
1 2 3 4
Curren 174 A 172 A 179 A 174 A
t
Voltage 19.8 V 19.8 V 19 V 20.0 V
Table 3.1: Parameter of various welding Power 3.4 kW 3.4kW 3.2kW 3.1kW
Weld 5mm/s 5mm/s 5mm/s 3.2 mm/s
speed
Wire 6.0m/min 6.4m/min 4.6m/min 6.5m/min
feed
Figure 3.1.1 : rate
Comparison of four Gas 100% 100% 100% 100%
welding seams with Argon Argon Argon Argon
various welding

The experimental started with WAAM part welding with 10 layers


and 3 string. At the first test we tried to weld the whole part without holding
time and the result shows very bad surface also it looks like the material is
flowing down. Then, for the next test, the surface was applied for two
minutes holding time for each layer and the entire geometry looks much
better. The temperature for the SS316L multiple layer welding process
shouldnt not be over 150°C because the effect of hot cracking.

Figure 3.1.2: Results of WAAM process by using normal


and holding time of 316L filler material
3.2 SIMULATION FEM

On this simulation, the all properties and parameter was follow on the actual
experimental 3 String 10 Layers on WAAM product. The parameter was uses
exactly same on the experimental which used the power heat input 3.4 kW.
Simulation also use a real size of the table, plate and weld bead in
millimeter(mm). Besides, the actual four clamping also assign on this
simulation. It was also including the 2 minute holding time at load case.

Figure 3.2.1: The Actual Heat source parameter used on the MSC Marc/Mentat.

In my case, the two simulation different temperature need to run on


the marc/mentat software. The normal and high temperature on the flow
curve need to determine their comparison result. So, it was necessary to
adjust some temperature at aflowmat of the MSC Mar./mentat. After through
some trial of the simulation, the thermal expansion is the main factor to run
the different temperature. So, thermal expansion for normal temperature are
2500K−1 and high temperature increase to 3000 K−1. Finally, both simulation
complete with show the right exit number which is 3004.

Figure 3.2.2: The result of 3 String 10 Layer simulation


In this simulation, the distortion result was determined by a certain
point on the specimen. The note depends on the total data needed for this
simulation. In this case, seven points were taken on the specimen's width and
the curve shows the distortion condition. After that, the MSC Marc Mentat
software clipboard can be copied to Excel and the graph can be interpreted
using mat-lab software. Therefore, for the purposes of comparison, both
normal and high temperature flow curve was used on mat lab software.

Figure 3.2.3: Distortion of normal Figure 3.2.4: Distortion of high


temperature simulation of WAAM. temperature simulation of WAAM.

Based on graph, the differentiate of the both simulation


approximately to 0 because the value only 0.0001. So, it was not having huge
impact to distortion at the high temperature of flow curve. Besides,
according the stress indicator also almost same to the both simulation. It also
happens on the equivalent von misses to the both simulation.

Next, the result simulation focus on the residual stress occurs on the
multi-bead layer of WAAM specimen. So, every 3 layer at every string was
taken on the simulation and the result was applying on the graph to show the
result however it was compressive or tensile stress. It simple to determine
both stress which are compressive stress shows negative (-) result and tensile
stress show positive result (+).

Figure 3.2.5: Appoint of nodes at every 3 layer of every string to determine


effective stress.
RESIDUAL STRESS 3S10L 3rd Layers
Stress Reliability Limit MPa 250
200 6th Layers

150 9th Layers

100
(+) Tensile
50 Residual stress
0
0 5 10 15 20 25 (-) Compressive
Residual stress
Width of weld filler (mm)
Figure 3.2.6: The graph of residual stress both simulation

Both simulation show the same result because of the differentiate


only 0.0001. So the results assume are same. The graph shows, only the
tensile stress (+) appear at WAAM simulation. In theoretically, the high
tensile stress may occur on the middle welding but on the multi-bead layer
of welding make the stress difficult to transform on that position.
Essentially, the effective stress at the based need determine
because WAAM fabrication as a product. So together the comparison, the
other simulation was created to measure of effective stress at base of
product without substrate and table whether effect or not.

Effective Stress at the Base Product


100
Stress Reliability Limit MPa

80
60
40 Figure 3.2.8: Simulation of WAAM
20 without substrate for a results

0
0 5 10 15 20 25
Width of weldfiller

Figure 3.2.7: result of effective stress at the base of WAAM


simulation
In fact, the stress occurs on the based WAAM product was not
have a lot stress compared to effective stress with complete cooling rate.
Although, the residual stress at the base suppose high stress according
multi-bead layer but it still below than residual stress occurs at that
complete cooling rate.
3.3 HARDNESS TEST

Hardness testing conducted with the Vickers hardness testing machines for
every layer of multiplayer welding by using Wire Arc Additive
Manufacturing(WAAM). The table shows, result of hardness test was
applied on the 10Layers 3String by Hardness Vickers machine.
Vickers Hardness Test (Diamond Indenter)
Load = 1 Kg

STRING 1 STRING 2 STRING 3


1S1L 161.3 1S1L 178.3 1S1L 166.8
1S2L 148.3 1S2L 158.8 1S2L 149.8
1S3L 188.8 1S3L 168.9 1S3L 142.9
1S4L 153.5 1S4L 181 1S4L 156.2
1S5L 187.5 1S5L 179.1 1S5L 160.5
1S6L 207.9 1S6L 195 1S6L 161
1S7L 218.1 1S7L 186.7 1S7L 170
1S8L 220.8 1S8L 175.6 1S8L 202.4
1S9L 208.4 1S9L 193.1 1S9L 177.4
1S10L 151.3 1S10L 163.7 1S10L 155.2

Table 3.3: Result of hardness test by Hardness Vickers machine.

String 1: - String 2: - String 3: -


Min: 148.3 HV Min: 158.8 HV Min: 142.9 HV
Max: 220.8 HV Max: 193.1 HV Max: 202.4 HV

Roughly, the result shows Vickers Hardness Test for every string
was approximately on each other’s. According to the result, string 3 shows
that have highest and lowest of Vickers hardness result compare to other
string but was not having a lot of different on each string. So this result
accepted to evaluate material's properties such as strength, ductility and
wear resistance, and helps to determine 316L material by refer on the data
type hardness.
3.4 X-Ray Diffraction (XRD)
The result of X-Ray Diffraction (XRD) show was different to the
residual stress occurs on the simulation. The process took 2 hours to
undergo for every node. Besides, Bragg diffraction angles need to set
up for the angle on finalize the residual stress. The result takes on
the simulation of the WAAM similarly node.

Nodes String String String


Table 3.4: Result of the residual stress for the Position 1 2 3
WAAM specimen. 3rd Layers
3237.56 540.97 796.91
6th Layers
197.16 -175.83 -697.78
9th Layers 2005.01
2487.21 -311.66
Figure 3.4.1: The nine-point
undergoing to the XRD

RESIDUAL STRESS 3S10L


3500 3237.56
3rd Layers

3000 2487.21 6th Layers


Stress Reliability Limit

2500 2005.01
9th Layers
2000
1500
796.91
1000 540.97
(+) Tensile
500 197.16
1 2 3 4 5
-175.83 Residual stress
0 -311.66
-697.78
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 (-) Compressive
-500
-1000 Residual stress
Strings
Figure 3.4.2: Result of the residual stress for the WAAM
specimen.

Based on the graph, the residual stress appropriately


different with result of effective stress by simulation. For the third
layer, there no have compressive stress but on the six and nine
layers have some compressive stress at the end of node. As we
know, the limitation of XRD machine can conduct only 6mm for
specimen. Therefore, the cutting process of the specimen might
effect to compressive residual stress.
3.5 Macrostructure

The macrostructure process conducted was to determine


similarity of the actual size by experimental and simulation
size of every weld bead. The dimension on millimeter for
each weld bead. By using the macrostructure machine, the
measuring was undergoing manually by used manual
method (ruler).

Figure 3.5.1: the manual method measurement by used ruler in millimeter (mm)

1 String (mm) 2 String (mm) 3 string (mm)


L H W L H W L H W
1 2 6 1 2 5 1 2 7
2 3 6 2 2 8 2 2 7
3 2 5.5 3 2 8 3 2 6
4 2 5 4 2 8 4 2.5 5
5 2 6 5 2 7 5 2 4
6 2 6 6 2 7 6 2 6
7 2 5.5 7 2 7 7 2 5
8 2 6 8 2 7 8 2 6
9 2 6 9 2 6 9 2 6
10 4 6 10 4 6 10 5 9

Table 3.5: the result of the all size weld filler on the WAAM specimen

2.5 mm

7.5 mm
Figure 3.5.2: Result of average of Figure 3.5.3: Result of size weld filler
weld filler (experimental) (Simulation)WAAM specimen.

3.6 Energy-dispersive X-ray spectroscopy(EDX)

Next, the process of EDX was conducted for the wire SS316L and
specimen. It was to observe the chemical composition of the results.
The six point randomly appoint at the wire SS316L surface, and
bottom, middle and top surface of specimen. The element form
periodic table which are C, Si, Mo, Cr, Mn, Fe, and Ni was called at
to every specimen result.

Bottom Middle Top Top

Figure 3.6.1: The surfaces of WAAM specimen applies EDX method.

Wire

Figure 3.6.2: The surfaces of Figure 3.6.3: The spot calculation of the chemical
wire SS316L specimen applies composition to specimen.
EDX method.

As the result, we can see the comparison of between catalog


SS316L and the actual product of SS316L which are wire and WAAM
product. On this method, the actual product need to emphasize about the
quantities carbon because it was main symbolic ‘L’ on the SS316L which
mean low carbon steel. Unfortunately, it was having large different about
carbon content at actual Stainless steel 316L with catalog guideline. It may
happen the exposure of the specimen and the specimen have not prepared
well.
CATA WIRE BOTTOM MIDDLE TOP
LOG
C 0.01 0. 0. 1. 1. 0. 0. 0 0. 0. 0. 0. 0. 1. 1. 1 0. 2. 3. 1. 5. 1 6. 1
K 4 7 1 1 8 7 2 2 2 1 2 7 1 6 5 9 7 4 3. 7 0.
1 7
Aver 0.8 0.15 1.38 6.916666667
age

Si 0.9 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1 0. 0. 0. 0. 0.
K 4 7 9 7 8 8 7 7 8 9 3 7 8 6 7 6 8 6 8 6 7 7

Aver 0.716666667 0.683333333 0.7 0.733333333


age

M 2.6 0. 1. 1. 1. 1. 1. 0. 0 0. 0. 0 0. 0. 0 0 0. 0 0. 0. 0. 0. 0. 0.
o 7 3 4 3 4 1 1 1 2 1 1 1 3 2 1 1 1 1
L
Aver 1.2 0.083333333 0.04 0.15
age

Cr 18.4 1 1 1 1 1 1 1 1 1 2 2 1 1 1 1 1 1 1 1 1 1 1 1
K 8. 7. 7. 7. 7. 7. 8. 8 7. 0 1. 8. 8. 8. 8. 9 8. 7. 9. 7. 6. 7. 7
4 9 6 4 7 4 8 9 1 7 6 8 6 3 6 2 5 4 5
Aver 17.73333333 19.08333333 18.66 17.53333333
age

M 1.8 2. 2. 2. 2. 2. 2. 2. 2. 2. 3. 4 2 2. 2. 2. 2. 2. 2. 2. 2. 2. 2. 2.
n 6 5 6 3 1 5 5 5 7 4 8 4 5 5 4 7 2 4 4 1 5
K
Aver 2.433333333 2.85 2.52 2.383333333
age

F Balan 6 6 6 6 6 6 6 6 6 6 6 6 6 6 5 6 5
e ce 8. 7. 7. 6. 6. 7. 5. 6. 7 7. 6. 2. 7 1. 8. 3. 9.
K 8 9 5 2 6 9 5 7 6 5 7 7 9 8 7
Aver 67.48333333 66.66 62.3
age

Ni 12.2 1 1 1 1 1 1 9. 1 1 9 7. 1 1 1 1 9. 9. 1 9 1 8. 8. 9.
K 2. 1. 2 2. 0. 2. 1 0. 0. 8 0. 0. 0. 0. 5 5 1. 2 6 9 3
8 6 2 6 3 7 9 4 5 3 1 8
Aver 11.91666667 9.65 9.98 9.933333333
age
Wire and WAAM 3 String 10 Layers specimen of 316L stainless steel chemical composition
4.0 Conclusions and Recommendations

In this study, a new technology improvement in the manufacture of


complex shaped components by used Wire Arc Additive Manufacturing
(WAAM) well established in industry. The quality of the product usually
was referring to the parameter used on the robotic welding and important
to verify the quality by visualize with some knowledge about welding
quality. An Austenitic Cr-Ni Stainless Steel 316L already used in WAAM
with excellent results, so it capable to proving the quality through
destructive test that have done. The size weld filler that have small scale
ratio differentiate between actual and simulation. Besides, specimen was
undergoing through the deformation residual stress after welding by using
X-RAY Diffraction (XRD). However, the limitation preparation and
advance technology at our lab cannot produce the result exactly to the
simulation FEM. Apart from waterjet cutting for specimen XRD method,
almost researcher suggest that use electroplating method to ensure the
stress of welding not disturb with others stress. Besides, the Energy-
dispersive X-ray spectroscopy(EDX) was done to the wire and WAAM
product but the element carbon to high on the specimen. Owing to some
factors as well as exposure of the product itself and maybe the lack of
preparation before apply EDX method. However, the result still has
similarities to certain main element such chromium, manganese and nickel.
In spite of that, the research will more detail and effectively when
conducted the experiment with high technology machine and tools. For
future research, if the experimental and simulation existed conducted
through right technique it was capable to expand the industry of additive
manufacturing since of the SS316L obviously usage on the marine and
aerospace industry which the high good superiority products.

5.0 REFERENCES
[1] S. W. Williams, F. Martina, A. C. Addison, J. Ding, G. Pardal, and
P.Colegrove, “Wire þ Arc Additive Manufacturing.”2015
[2]. Rapid Prototyping, and Direct Digital Manufacturing, 2nd ed.,
Springer, New York, 2015.
[3] S. W. Williams, F. Martina, A. C. Addison, J. Ding, G. Pardal, and P.
Colegrove, “Wire þ Arc Additive Manufacturing.” 2010
[4] R. K. Desu, H. Nitin Krishnamurthy, A. Balu, A. K. Gupta, and S. K.
Singh, “Mechanical properties of Austenitic Stainless Steel 304L and
316L at elevated temperatures,” J. Mater. Res. Technol., vol. 5, no. 1,
pp. 13–20, Jan. 2016.
[5] Magnus Eriksson,*, 1, Malin Lervåg2 , Camilla Sørensen2 , Andreas

duplex stainless steel using WAAM” MATEC Web of Conferences


188, 03014 (2018)
Robertstad2 , Bård M. Brønstad2, “Additive manufacture of super

You might also like