Downloaded from https://iranpaper.

ir
https://www.tarjomano.com https://www.tarjomano.com

Proceedings of ASME Turbo Expo 2023

Turbomachinery Technical Conference and Exposition
GT2023
June 26-30, 2023, Boston, Massachusetts, USA

GT2023-103043

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

DEVELOPMENT OF ADDITIVE MANUFACTURING GAS TURBINE HOT GAS PATH VANES
AT BAKER HUGHES

Massimiliano Cecconi Iacopo Giovannetti

Baker Hughes Baker Hughes
Florence, Italy Florence, Italy

ABSTRACT and monitored by periodic inspections. This allows to complete

This paper describes the efforts, the lessons learned and the the durability assessment in real service environment.
success that led, over a period of 5 years, to the development, The work is completed with an overview on the industrialization
validation and finally industrialization of high-pressure turbine challenges of the technology. It has been provided guidelines in
vanes manufactured with additive Direct Metal Laser Melting components design and printing strategy to mitigate macro
(DMLM) technology. cracks formation on high gamma prime content material,
The introduction of the study collects the requirements for gas including post-print processes as EDM cut removal, thermal and
surface treatments.
turbine vanes in the range of F class and identify the material's
grade and chemical composition of the alloys. As well it is It is finally compared the CO2 footprint of additive technology
summarized the conflict among the processability and desirable versus investment casting, demonstrating a significative
mechanical properties of Nickel base alloys. reduction in the CO2 emission in gas turbine vanes
manufacturing process using additive technology.
The paper explains the procedure used to simultaneously
optimize the alloy and the DMLM process parameters for a
Keywords: Additive Manufacturing, New Materials, Gas
Nickel based superalloy with high gamma prime content to
Turbine Vanes
manufacture complex geometries as high-pressure vanes in the
turbine hot section. Metallurgical defects formation in
additively manufacturing superalloys are strongly linked to
process parameters, such as laser power, scan speed, layer NOMENCLATURE
thickness, scan strategy, etc. but are also directly linked to the GT Gas Turbine
specific geometry of the component to be processed. DMLM Direct Metal Laser Melting
Contouring parameters as downskin/upskin that have higher EDM Electro Discharge Machining
influence in the residual stresses of the parts and lead to a RT Room Temperature
mechanical properties debit to fatigue resistance. °C Celsius degrees
A “fast-work” approach was used starting with a limited Al Aluminum
validation of the components in a back-to-back comparison Ti Titanium
with the traditional technology in lab environment. Over the IC Investment Casting
standard mechanical testing samples, it is introduced subsize AM Additive Manufacturing
component mechanical testing and full-scale component SEQV Von Mises Equivalent Stress
thermal cycle above 1’000°C. DPI Dye Penetrant Inspection
MFG Manufacturing
A second step to mature the technology has been to produce and
install a first set of vanes and shrouds to be tested in a real
validation engine to verify resistance to the main failure modes,
in particular cyclic type of damage as low cycle fatigue and
crack propagation.
In the paper the final validation step is also presented. In
this phase the components are installed in a production engine

1 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

1. INTRODUCTION
Additive Manufacturing has proved to be a very effective
technology when applied to parts or components that have a
sophisticated design and that, for the peculiarity of their

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

application, do not need to be produced in a very large scale. Its
capability to significantly reduce the lead time of the supply
chain is of extreme importance when introducing a new design,
especially when multiple concepts need to be verified.
The abovementioned characteristics make the additive
particularly suitable to produce hot gas path components for gas
turbines because of their design and supply chain complexity. In
fact, when compared to more traditional technologies like
investment casting, the additive brings the benefit to
significantly simplify the supply chain, with most of the
transformation processes happening under a single rooftop. This
leads to economical benefits, supply chain management FIGURE 01: High temperature gas turbine superalloys [2]
simplification and a better control over market globalization
instabilities. Logistics and transportation costs can benefit of this The selected alloy family, having a gamma prime content in
supply chain simplification as far as the process quality can be the order of 60%, is non-weldable (see figure 02) [5], [6] using a
better kept under control. powder bed fusion process. In order to make the alloy more
The steps that Baker Hughes went through to develop and processable by AM, the need to modify its chemistry came out
validate Gas Turbine high pressure vanes manufactured using pretty soon to mitigate the susceptibility to crack [1]. At the same
Direct Metal Laser Melting as additive technology is reviewed. time the chemistry optimization didn’t affect the Al and Ti
This works starts from the alloy development and content which are critical elements to achieve the desired gamma
characterization of a high gamma prime nickel-based superalloy prime content and the required mechanical properties at
that is described in detail in [1]. The application to gas turbine temperatures higher than 900°C.
vanes required the optimization of the manufacturing additive Immediately after the first printing trials, it was pretty clear
parameters and the alloy chemistry optimization. The required that the alloy chemistry optimization only wasn’t enough to
target is to obtain parts defect free capable to reach the expected manage the material sensitivity to generate cracks and a AM
life and durability. parameter optimization was necessary. This approach permitted
to keep the size and dimension of the cracks to values that could
be easily healed by the hot isostatic pressing post process.
2. MATERIALS AND METHODS
The development of the gas turbine vanes has required a
dual effort for the base material and for the manufacturing
process. The development of the gas turbine components has
required a dual effort for the base material and for the
manufacturing process.
The development of the gas turbine components has
required a dual effort for the base material and for the
manufacturing process. While for the material the most logic
choice has been to develop from a well-known casting alloy (see
figure 01), like the CM247LC or René108 [2], [3] [4], the
additive process, including the complex post process steps,
required a bunch of innovative approaches. The focus of the
study was only on equiaxial superalloy, directionally solidified
or single crystal alloys were excluded from the investigation
scope because of the AM process constraints.
This choice allows to target the component life requirements FIGURE 02: Weldability diagram of Nickel base superalloys [6]
for the application to gas turbine first stage vanes in Class F Gas
Turbine. The validation of the manufactured components followed 3
main steps. The first one was performed at lab scale aiming to
compare the low cycle fatigue performance of additively
manufactured vanes with respect to the same component made
by investment casting. The parts were cycled in a furnace
simulating very strong thermal cycles with heating and rapid

2 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

cooling (par. 3.1). Early failures drove an alloy and parameters Microcrack and porosity defects were managed by alloy
first optimization. chemical optimization in combination to AM process parameters
Having passed this first step of validation, a second as previously reported [1]. To validate the material properties in
validation phase was directly performed on a NovaLT® additive manufacturing components, a valuable approach is to

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

validation engine, where some of first stage vanes made in use samples as similar as possible to the final component. The
additive were installed in a rainbow configuration with other mechanical behavior of GT vanes thin walls is better represented
produced by investment casting. The engine was operated at full by flat specimens with respect to cylindrical. For example, the
load including transient cycles to simulate the most critical real figure 05 shows a significative difference in elongation at high
operating conditions for vane durability. temperature between the two kinds of specimens.
The test results suggested further optimization of the AM In order to better simulate the material properties a
process that led to the production of a third generation of vanes further step is the use of custom samples sampled directly from
that were finally installed in a production engine. This engine has the printed component.
today cumulated more than 20’000 hours of successful operation
Tensile Elongation comparison
with numerous start up and shut down.
100%
3. RESULTS AND DISCUSSION 90% Additive - cylindrical

Normalized relative value

80% Additive - flat
70%
3.1 Lab scale iterations
60%
The typical defects generated by AM process of a high
50%
gamma prime nickel-based superalloy can be typically classified
40%
as:
30%
20%
 Microcracks (figure 03) 10%
 Porosity (figure 04) 0%
 Macrocracks (figure 06) RT 871°C
 Roughness (figure 09) FIGURE 05: Normalized tensile properties comparison on flat
 Dimensional vs cylindrical samples

From the first printing trials it is noted a macrocrack

formation in the vanes (see figure 06). The cause of this defect
was linked to the combination of a high value of residual stress
and “poor” weldability of the alloy, driven by its gamma-prime
phase content. The cracks generally occurred during the heat
treatment post-process when additive process residual stress is
released.

FIGURE 03: Typical microcrack defect processing high gamma

prime content alloys. Cross sections observed by means of optical
microscope of the material structure with microcracking defects

FIGURE 06: Typical macrocrack defect in GT additive vanes

The residual stresses were kept under control through a

design for additive approach leveraging process simulation. The
optimization involved critical component design features like
fillet radii as well as the AM printing strategy like supports
design and part orientation on the construction platform (see
figure 07).
FIGURE 04: Typical porosity defect processing high gamma
prime content alloys. Process porosity observed by means of dye
penetrant inspection (picture on the left side) and in a metallography
cross section by means of optical microscope (picture on the right side)

3 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

FIGURE 07: Examples of support design on GT high pressure
vanes having macrocracking issue

Residual stresses were managed also by a design of the heat

treatment, introducing a dedicated stress relief process. A
difference between the residual stress in additive manufacturing FIGURE 09: Fatigue debit due to roughness: “machined” is the
vane “as built” condition and after heat treatment is represented standard sample while “as built” is the sample having a rough surface
in figure 08. obtained directly by AM process. The image shows a detail of a crack
on a GT vane airfoil geometry which is demonstrated to start from
internal “as built” surface instead of the external machined surface.
Additive process residual stress comparison
1200 Specific surface finishing technologies were developed in
1000 SEQV as built order to recover the fatigue debit in critical component locations.
Figure 10 represents the comparison between investment
Stress [MPa]

800 SEQV after heat treatment

casting, additive “as built” and additive with finishing process
600 applied.
400
200 Roughness comparison
0 30
0 0,2 0,4 0,6 0,8 1 Ra
Depth [mm] 25
Rz
20
FIGURE 08: Residual stress on additive manufactured GT vane
μm

15
in “As built” and heat-treated conditions
10

5
The component shrinkage and deformation generated by the
0
residual stress release, can generate significative geometrical out Additive (as built) Additive (finished) Investment Casting
of tolerance. This is not only potentially impacting aerodynamic
performances but also the cooling of the vane because of out of
tolerance holes size and shape. For these reasons, a morphing FIGURE 10: Surface roughness comparison of GT vane in
process (i.e. iterating between real component dimensions and different finishing conditions
3D model) was applied to the original vane geometry allowing
to target the required dimension. 3.2 Lab scale validation
The first step of the validation process was aimed to
AM process typically generates high levels of surface evaluate, in a comparative way, the GT vane additive solution
roughness that results in fatigue life reduction of the versus the equivalent investment casting one. The most
components. For example, figure 09 shows the low cycle fatigue demanding load for a double airfoil GT vane in operation is the
results obtained on the same component in its “as built” and after thermal cycling at high temperature that drives low cycle fatigue
machining conditions. damage.
The target cycle test was set to 30 cycles. A sequence of
thermal cycles has been performed by instantaneously inserting
two components in an air furnace at 1000°C, removing them and
cooling down by forced air convection. Every 5 cycles the vanes
were inspected by means of DPI to detect cracks presence. The
figure 11 show the two components layout during forced cooling
and a typical temperature trend on vane platform during cooling.

4 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

FIGURE 11: Thermal Stress Test on GT Vanes. Cooling Layout
(left) and vane platform temperature trend during cooling (right)

The test was performed on a stage1 high pressure vane of a

NovaLT® gas turbine made in investment casting and on the
same component made by additive manufacturing. Investment FIGURE 13: Additive vanes installation for a rainbow test on
casting alloy was the CM247 LC equiaxed alloy, while the NovaLT® gas turbine
additive manufacturing vane was made in 2 different alloy
chemistries: CM247 LC alloy processed by additive and NP110, The engine test performed an overall of 56 cycles and 81
a proprietary alloy patented by Baker Hughes (EP3426811B1) firing hours. The cycles consisted of a ramp from zero to full
[1]. The test, that was repeated two times on the NP110 alloy load, permanence at full load for engine thermal stabilization and
brand new condition, was performed on investment casting parts finally emergency shut down. The engine was re-started after its
in brand new condition as well as after service condition. complete cool down. This cycling was determined as the most
The results in figure 12 show that components in brand new demanding for the first stage vane because of the highest
condition made of additive NP110 alloy and investment casting generated thermal stress. This test has been able to put in
successfully passed the test. The investment casting component evidence eventual low-cycle fatigue limits of AM parts.
in serviced condition failed the test after few cycles as well as Vanes were disassembled, cleaned and visually inspected.
the additive CM247 LC alloy. No defects as plugged cooling holes or other kind of damages
35 (cracks) were detected by visual examination (see figure 14)
30
Target test cycles neither by DPI (figure 15).
25 # cycles without cracks
20
15
10
5
0
Investment Casting Investment casting Additive CM247 LC Additive NP110 -
- Brand new - Serviced - Brand new Brand new FIGURE 14: Visual inspection of additive vane after validation
test.
FIGURE 12: Thermal Stress Test results on GT vanes in
different conditions.

This test, even if not fully representative of real engine

operation, represents a good method to assess in a comparative
and quick way the different process solutions, mitigating
component early failures before field validation test.

3.3 Validation engine test

The second validation step was performed on the NovaLT®
gas turbine validation engine.
A first campaign was conducted by means of a rainbow test
with 5 additive vanes (double airfoil) and 17 investment casting
vanes having the same geometry have been assembled in the FIGURE 15: Dye Penetrant Inspection of serviced additive vane
same 1st high pressure stage of the validation engine (figure 13). (validation test). No indications detected.

5 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

The components were inspected also by means of subject to a periodic boroscopic inspection to check their
metallographic investigation. A detail is reported in figure 16: integrity. As example figure 18 reports the result of two
after service sound material condition is observed. boroscopic inspections.

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

FIGURE 16: Micrographic examination detail of serviced
additive vane (validation test). Detail of the cut up of an airfoil (central
picture) and the typical appearance of the material in the cross sections.
It is observed sound material conditions in thin trailing edge section on
the left, thicker section on the right FIGURE 18: Boroscopic inspections of additively manufactured
vanes after 4’000 hours (left) and 15’000 hours (right) on NovaLT® gas
The positive results collected at the end of the first turbine. It is possible to observe the trailing edge portion of additive GT
campaign, encouraged the execution of a second validation vanes. No defects have been detected.
campaign focused on the crack propagation. The same
NovaLT® validation engine was used. For this campaign five Once the gas turbine “fleet leader” reached 20’000 hours
additive vanes were used, two of them had an artificially and 126 fired starts, the vanes were disassembled and inspected
manufactured defect. The other 17 vanes were produced by by means of visual inspection, DPI and X-Ray tomography scan.
investment casting, two of them having the same artificially Visual inspection has not revealed defects (figure 19),
manufactured defect. without noticeable appearance differences between AM and IC
The components were subject to the same kind of cycles of components. The AM vane airfoil leading edge shows an even
the first campaign, collecting overall 67 cycles and 100 firing better performance than the equivalent investment casting one
hours. The same kind of inspections have been performed after for what concerns the oxidation damage.
the disassembly on all the parts: visual inspection by means of
stereoscope microscopy and DPI.

FIGURE 17: Inspection of additive vane (validation engine test).

On the left picture it is reported the detail of artificial defect inspected
by dye penetrant. On the right picture the stereoscopic inspection detail.
No defect propagation is detected.

As shown in figure 17, no propagation of the artificial

defects has been revealed by the inspections carried out at the
end of this second test campaign in AM and IC components.

3.4 Production engine in field validation

The excellent results collected during the two campaigns on FIGURE 19: Additively manufactured and investment casting
the validation engine unlocked the possibility to proceed with a vanes after 20’000 hours of field operation in the same NovaLT®
durability test on production engines for in field operation engine
validation.
Additive vanes were installed on 3 different GT production No indications were detected by DPI (figure 20) and by X-
engines (family of Baker Hughes NovaLT®). The same rainbow Ray computer tomography inspection (figure 21)
approach was used installing in each engine only a reduced
number of AM vanes, typically five additive components have
been installed, the remaining were in IC. The components were

6 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

FIGURE 20: Dye Penetrant Inspection of additively
manufactured vane after 20’000 hours. No indications detected.
FIGURE 22: Design for additive example (enhanced fillet radii)

FIGURE 21: Computer tomography of an additively

manufactured vane after 20’000 hours. No indications detected.

The components inspected at 20’000 hours have been

reinstalled into the engine to cumulate additional service hours FIGURE 23: Design for additive example (avoid bumps on
and to continue the durability validation. internal cavities)

3.5 Manufacturability and industrialization The part removal from the construction plate by EDM must
The industrialization phase of the AM GT vanes suggested be the first post-process to be performed before any other thermal
some key learnings. or surface treatment. Stress relief heat treatment and hot isostatic
Double or triplet vane design are more prone to macro pressing must be optimized to guarantee material integrity and
cracks formation, singlet vane design solution is preferrable reduce macro and micro cracking.
when possible. Surface treatment like shot peening are recommended for
The optimization of laser parameters directly on the most critical locations of the vane to avoid macrocracking.
components instead of samples mitigates the risk of internal Its optimization (Almen intensity, etc.) is key to avoid
material integrity. By this method it is possible to validate the component deformation due to reduced thickness of some of its
material properties using “custom” samples. areas or to avoid internal defects induced by the post-process.
Optimization of the fillet radii is more critical than other The roughness of the as-built vane may lead to a couple of
manufacturing processes (Figure 22 and 23) because of the AM undesired consequences. The aerodynamic performance of the
residual stresses. airfoil can be affected by the increased vorticity in the boundary
Cooling holes dimension shrinkage is a function of printing layer of the gas path and the cooling features (holes, etc.) may
orientation and features. It is important to apply a morphing change their consolidated performances. To target design
process to create a transfer function for small size holes requirements finishing post processes are needed. Two different
considering the manufacturing process capability. super-finishing processes were adopted: one for the external
Process induced residual stresses require an optimization of portion of the vanes and one for the internal cavity. Due to its
the printing orientation and the addition of overstock material in limited accessibility, the last one required a specific process
critical locations to avoid macro crack formation after heat development to achieve the required roughness values to avoid
treatment. any possible fatigue debit (see figure 24).
Trailing edge thickness should be kept at the maximum
possible value permitted by the range of tolerances for the
additive process, taking properly into account the effect of the
superfinishing post process. To find a balance is often not simple
because the airfoil trailing edge is one of the most critical areas
from structural point of view because of its limited thickness
required by aerodynamic performance.

FIGURE 24: Additive GT vane with internal and external

surfaces finishing post-process

7 Copyright © 2023 by ASME and Baker Hughes

 Downloaded from https://iranpaper.ir
https://www.tarjomano.com https://www.tarjomano.com

3.6 CO2 Footprint Comparison ACKNOWLEDGEMENTS

AM technology is playing an important role also in the The authors thank Baker Hughes for financial support and
decarbonization targets of the supply chain operations. An all the collaborators that have provided their contribution to this
evaluation of the CO2 emissions in the manufacturing cycles of work.

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2023/87066/V11BT23A005/7045314/v11bt23a005-gt2023-103043.pdf by University of Edinburgh user on 12 November 2023

GT vanes has been performed and presented elsewhere [7]. The
study shows that in the case study of GT vanes production, the REFERENCES
additive process can reduce the CO2 generated more than 15% [1] Callisti M., Dimatteo A., Development of Nickel-based
(see figure 25) with respect to the traditional investment casting. superalloys for hot gas path components via additive
manufacturing, ASME TURBOEXPO paper GT2023-102874,
Investment Casting Additive Manufacturing 2023
100%
Normalized kg CO2 equivalent produced

90%
[2] Jin T., Zhou J., Wang X., Liu J., Sun X., Hu Z., Research
80% process on microstructural stability and mechanical behavior of
Heat
Heattreatment & Coating
treatment & coating
70%
advanced Ni-Based single crystal superalloys, Acta
60%

50%
Post
Postprocess
process &
&Machining
machining Metallurgica Sinica, Vol. 51, Issue 10 (1153-1162), 2015
40% Raw
Rawmaterial
materialproduction
production [3] Joseph N. Ghoussoub et Al., On the Influence of Alloy
30%
Composition on the Additive Manufacturability of Ni-Based
20%
10%
Superalloys, METALLURGICAL AND MATERIALS
0% TRANSACTIONS A, VOLUME 53A, March 2022.
Investment casting
FIGURE 25: Comparison of Life Cycle Environmental Impact Additive MFG
[4] Salome Sanchez et al., Powder Bed Fusion of nickel-
for the production of a GT vane ring made by investment casting process based superalloys: A review, International Journal of Machine
versus additive manufacturing one Tools & Manufacture 165 (2021) 103729.
[5] Md.Shahwaz, Prekshya Nath, Indrani Sen, A critical
4. CONCLUSION review on the microstructure and mechanical properties
The development of high-pressure GT vanes by additive correlation of additively manufactured nickel-based superalloys,
manufacturing has required and important effort in Baker Journal of Alloys and Compounds 907 (2022) 164530.
Hughes along several years and a synergic approach through the [6] Basak A., Additive Manufacturing of High-Gamma
leverage of multiple competencies in material science, Prime Nickel-Based Superalloys through Selective Laser
manufacturing and testing engineering. The program developed Melting (SLM), Materials Science, 2019.
a key knowledge that can be summarized as follows. [7] Serra A., Fantaccione L., Fantozzi F., Bartocci P., Buia
When dealing with nickel-based superalloys having a high G., Gandini S., Effects on sustainability by Redesign of a 1° stage
content of gamma prime, the chemistry and the heat treatment gas turbine nozzle ring from Investment Casting to Additive
must be optimized for the additive process. This is the main Manufacturing: A Life Cycle Environmental Impact comparison,
reason why we decided to patent an original alloy, namely the ASME TURBOEXPO paper GT2022-78256, 2022
NP110, and develop specific process parameters for it.
The development of the alloy and of the additive process
parameters cannot be in series but must proceed in parallel with
multiple improvement iterations. The process parameters that are
allowed for an alloy like the NP110 typically seat in a narrow
window [1].
All the validation tests, also at lab scale, must be as much as
possible representative of the potential failure mode in field to
quickly address the weakness of the alloy and of the process.
The large amount of experience collected during the tests of
the vanes installed in the validation and production engines has
proven the validity of the additive manufacturing as a solid
technology for this challenging class of alloys.
The use of AM for high pressure gas turbine vanes is today
a valid alternative to the more consolidated investment casting
process. It is especially true for not large production series and
when a fast re-design is needed; it has also the additional benefit
to protect the supply chain from uncontrollable instabilities and
to address the CO2 reduction targets.

8 Copyright © 2023 by ASME and Baker Hughes