Field of the Invention

This disclosure relates to methods and systems for digital design and qualification of products. The methods and systems according to the disclosure integrate specific knowledge about design, materials, manufacturing and performance to improve the development process of a part. The methods and systems according to the disclosure are specifically adapted for selecting materials for manufacturing a part using additive manufacturing processes.

Background of the invention

This disclosure relates to developing an end-to-end digital part design and qualification platform by integrating intelligent solutions for each key step in the product development. These keysteps include design, material selection, manufacturing, and performance assessment. To enable this, a seamless bi-directional communication of all mentioned technologies is developed on an integrated platform.

When selecting an adequate material for manufacturing a part, it is known to use a database comprising a list of materials and key parameters relating to each of the materials.

During the design process for a part to be manufactured, a computer model of the part should be obtained. This computer model is used, in combination with load data relating to the estimated load on the part, to determine the stress levels in the part induced by the estimated load on the structure. In a next step, high stress areas can be identified in the part. The stress level in these high stress areas can be compared with determined threshold values for various materials to assess which material would be best adapted for manufacturing the part.

In practice, the process for selecting an adequate material for manufacturing a part is an iterative process.

The specific design of a part to be manufactured, will have an effect on the functional performance of the manufactured part. However, in existing methods and systems for developing a part, specific knowledge relating to the actual design of the part to be manufactured and its manufacturability, is not or not sufficiently included in the process for selecting an appropriate material. Moreover, the actual manufacturing process that is selected for obtaining the part will have an effect on the design and consequently the functional performance of the manufactured part. However, in existing methods and systems for developing a part, specific knowledge relating to the actual manufacturing process is not or not sufficiently included in the part development process. These observations are particularly true for parts that are obtained using additive manufacturing processes.

For the reasons mentioned above, the development process for a part, including material selection and performance assessment, can be both time consuming and costly. Lengthy and costly design optimization, material selection and performance assesment processes will have many negative effects on the total time needed for development of the part.

In view of the observations above, it appears that there is need for an improved method and system for developing a part, including optimizing design for manufacturing and cost, and at the same time optimizing the selection of an appropriate material for best performance of said part, that can avoid or at least reduce the disadvantages of the mentioned methods and systems described above.

Summary of the invention

Various examples of the technology disclosed herein provide methods and systems for integrating the development process for a part. Further examples provide methods and systems for creating design, material and manufacturing related databases and integrating the knowledge stored in those databases to thereby improve the development process for a part.

For instance, during the design process for a part to be manufactured, a computer model of the part can be obtained either by taking advantage of and prescreening the existing CAD database with an assigned material or by generating a new model using the known spatial and static structural loading constraints, and by taking advantage of a manufacturing method and cost optimization based on the material selection. This computer model is used, in combination with the complete static and dynamic load data to be applied on the part, to determine the induced stress levels in the part. In a next step, high stress areas can be identified in the part. The stress level in these high stress areas can be compared with determined threshold values for materials stored in a materials database to assess which material would be best adapted for manufacturing the part.

In some examples a method for integrating the development process for a part, comprises:

- creating a design database, comprising designs of parts, wherein the design database comprises for each of the designs key parameters relating to the performance of the part; - receiving the requirements of the part to be developed, including key parameters relating to the required performance of the part;

- comparing the key parameters relating to the required performance of the part to be developed with the key parameters relating to the performance of parts stored in the design database to evaluate the suitability of the designs in the design database to meet the requirements of the part to be developed,

- identify the nearest available design in the design database based on the suitability to meet the requirements of the part to be developed, and

- if the nearest available design meets all requirements of the part to be developed, selecting said nearest available design as the initial part design, or

- if the nearest available design does not meet all requirements of the part to be developed, improving said nearest available design to obtain an initial part design that meets all requirements of the part to be developed.

By way of a further example, the designs in the design database are sorted in a number of product families, wherein each of the designs in a product family shares a determined number of design features with the other design in the same product family.

Additionally, in further examples the method comprises:

- storing in the design database parametric CAD models for each of the designs,

- generating a verified parametric CAD design for the part to be developed and

- comparing the verified parametric CAD design of the part to be developed with the parametric CAD design of the designs stored in the design database, to thereby identify the nearest available design.

In other examples the method comprises:

- using the initial part design for computational part performance analysis,

- identifying by means of the computational part performance analysis critical areas, indicative for performance parameters, such as maximum stress, fatique life and pressure drop, in the initial design of the part to be developed, and

- improving the initial part design using the results of the computational part performance analysis to thereby obtain an improved initial design for the part to be developed.

By way of example the method comprises:

- creating a materials database, comprising a number of materials, wherein the materials database comprises for each of the materials key parameters relating to the performance of the material; - using the initial part design for computational part performance analysis,

- identifying by means of the computational part performance analysis maximum stress areas in the initial design of the part to be developed, and

- comparing the results of the computational part performance analysis with the key parameters relating to the performance of the material, for the materials stored in the materials database, and

- selecting a material adapted for the part to be developed based on said comparison between the results of the computational part performance analysis of "as manufactured" part for critical to quality performance requirements in the initial design of the part to be used with the key parameters relating to the performance of the material, for the materials stored in the materials database.

In some examples, the method further comprises:

- creating a manufacturing database, comprising for a number of manufacturing processes key parameters relating to materials for which each of said manufacturing processes is adapted and key parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured,

- selecting a manufacturing process adapted for processing the selected material to manufacture the part to be developed,

- using the key parameters of the selected parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured to improve the initial part design and to thereby obtain an improved initial design for the part to be developed.

In some examples, the method further comprises:

- evaluating several manufacturing processes with different material combinations for the part to be developed,

- generating key parameters relating to the influence of the manufacturing process on the part to be developed, and

- updating the intelligent manufacturing database by adding said generated key parameters during the manufacturing of the part to be developed.

By way of example the selected manufacturing process is an additive manufacturing (AM) process, the method further comprising:

- creating an integrated quality assurance module comprising data relating to the manufacturing of a part to be developed, and wherein the step of generating key parameters relating to the influence of the manufacturing process comprises: - obtaining key parameters of the additive manufacturing process during the manufacturing of the part, and

- adding the obtained key parameters of the additive manufacturing process during the manufacturing of the part to the integrated quality assurance module.

In additional examples the method comprises:

- creating a manufacturing database, comprising for a number of manufacturing processes key parameters relating to materials for which each of said manufacturing processes is adapted and key parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured,

- selecting a material for manufacturing the part to be developed,

- selecting a manufacturing process adapted for processing the selected material to manufacture the part to be developed,

- using the key parameters of the selected parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured to improve the initial part design and to thereby obtain an improved initial design for the part to be developed.

In some examples, the method further comprises:

- manufacturing the part to be developed by means of the selected manufacturing process,

- generating key parameters relating to the influence of the manufacturing process on the part to be developed, and

- updating the intelligent manufacturing database by adding said generated key parameters during the manufacturing of the part to be developed.

By way of example the system for integrating the development process for a part, comprises:

- a design database, comprising designs of parts, wherein the design database comprises for each of the designs key parameters relating to the performance of the part;

- a processor comprising a communication interface to communicate with an input device configured to forward requirements of a part to be developed to the processor, wherein the processor comprises a communication interface to communicate with the design data base, the processor being configured to compare the key parameters relating to the required performance of the part to be developed with the key parameters relating to the performance of parts stored in the design database to evaluate the suitability of the designs in the design database to meet the requirements of the part to be developed and to identify the nearest available design in the design database based on the suitability to meet the requirements of the part to be developed, wherein the processor is further configured, if the nearest available design meets all requirements of the part to be developed, to select said nearest available design as the initial part design, or , if the nearest available design does not meet all requirements of the part to be developed, to improve said nearest available design to obtain an initial part design that meets all requirements of the part to be developed.

By way of further example the system further comprises:

- a materials database, comprising a number of materials, wherein the materials database comprises for each of the materials key parameters relating to the performance of the material, wherein the processor comprises a communication interface to communicate with the materials database, the processor being configured to use the initial part design for computational part performance analysis to thereby identify by means of the computational part performance analysis critical areas, in view of static and fatigue requirements, in the initial design of the part to be developed, and to compare the results of the computational part performance analysis with the key parameters relating to the performance of the material, for the materials stored in the materials database, and the processor further being configured to select a material adapted for the part to be developed based on said comparison between the results of the computational part performance analysis with the key parameters relating to the performance of the material, for the materials stored in the materials database.

Additionally, in further examples, the system according comprises:

- a manufacturing database, comprising for a number of manufacturing processes key parameters relating to materials for which each of said manufacturing processes is adapted and key parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured, wherein the processor comprises a communication interface to communicate with the manufacturing database, the processor being configured to select a manufacturing process adapted for processing a selected material to manufacture the part to be developed and to use key parameters of the selected manufacturing process relating to the influence of the manufacturing process on the performance of the part to be manufactured to improve the initial part design and to thereby obtain an improved initial design for the part to be developed.

Additionally, in further examples the method comprises:

- creating a materials database, comprising a number of materials, wherein the materials database comprises for each of the materials key parameters relating to the performance of the material;

- receiving the requirements of the part to be developed, including key parameters relating to the required performance of the part;

- obtaining an initial design for the part to be developed; - performing performance analysis using the key parameters relating to the required performance of the part to be developed to thereby identify stress zones in the part to be developed;

- comparing the stress levels in the identified stress zones with the key parameters relating to the performance of the materials stored in the materials database to evaluate the suitability of the materials in the materials database to meet the requirements of the part to be developed, and

- selecting a material adapted for the part to be developed based on said comparison between the key parameters relating to the required performance of the part to be developed and the key parameters relating to the performance of the materials stored in the materials database.

In other examples, the step of creating a material database comprises:

- obtaining key parameters relating to the performance of a least a first material by using experimentation and

- obtaining key parameters relating to the performance of a least a second material by using means of a computational process.

In some examples, the method further comprises:

- after the step of selecting a material for the product to be developed, improving the initial design by using the key parameter.

In other examples, the method further comprises:

- creating an intelligent manufacturing database, comprising for a number of manufacturing processes key parameters relating to materials for which each of said manufacturing processes is adapted and key parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured, and after the step of selecting a material for the part to be developed,

- selecting a manufacturing process adapted for processing the selected material to manufacture the part to be developed,

- using the key parameters of the selected parameters relating to the influence of each of the manufacturing processes on the performance of the part to be manufactured to improve the initial part design and to thereby obtain an improved initial design for the part to be developed.

Additionally, in other examples, the method further comprises:

- manufacturing the part to be developed by means of the selected manufacturing process, - generating key parameters relating to the influence of the manufacturing process on the part to be developed, and

- updating the intelligent manufacturing database by adding said generated key parameters during the manufacturing of the part to be developed.

In other examples, the selected manufacturing process is an additive manufacturing (AM) process, the method further comprising:

- creating an integrated quality assurance module comprising data relating to the manufacturing of a part to be developed, and wherein the step of generating key parameters relating to the influence of the manufacturing process comprises:

- obtaining key parameters of the additive manufacturing process during the manufacturing of the part, and

- adding the obtained key parameters of the additive manufacturing process during the manufacturing of the part to the integrated quality assurance module.

In order examples, the method further comprises dividing the volume of the part to be developed into different stress zones, wherein optimum print parameters are identified for said different stress zones, in order to meet the structural requirements for each of the stress zones, to thereby increase the productivity during printing of the part to be developed.

In other examples, wherein the selected manufacturing process is an additive manufacturing (AM) process, the method further comprises using a powder melting model to predict melt- pool attributes, such as the width, depth, shape and length of the melt-pool, and a transient temperature profile depicting temperature gradient and cooling rates.

In other examples, wherein the selected manufacturing process is an additive manufacturing (AM) process, the method further comprises using solidification and solid-state transformation modelling, to predict solidification microstructure morphology, such as micro-segregation, dendrite size, dendrite shape, dendrite arm spacing, phases, grain type and grain size, for different laser heat energy input levels.

In other examples, wherein the selected manufacturing process is an additive manufacturing (AM) process, the method further comprises using homogenization modelling, to predict the mechanical properties at as-built and heat treatment conditions. Brief Description of the Drawings

Fig. 1 schematically shows the workflow of a conventional development process for a part; Fig. 2 schemattically shows a holistic approach for a digital integrated workflow for the development of a part;

Figs. 3-6 show different integration steps which are used to allow improvement of the part development process;

Fig. 7 shows a workflow relating to integrated manufacturing intelligence, which is adapted for robust additive part design and manufacturing;

Fig. 8 represents part of the workflow of Fig. 7 and shows in more detail the interaction between an additive manufacturing device and an Integrated Quality Assurance Module;

Fig. 9 shows a workflow wherein a step of pre-screening of existing designs is used to improve and accelerate part development;

Fig. 10 represents the process of continuously improving materials intelligence using artificial intelligence and platform integration;

Fig. 11 represents a workflow relating to material selection and a related workflow relating to material qualification, and

Fig 12. shows a workflow relating to material development.

Description of preferred embodiments

In the present disclosure, the term ‘functional performance’ is used to indicate the capability of a part to accomplish a certain task. This means that the functional performance gives an indication of the amount of stress, torque or any other functional load that can be applied on the part, without failure of the part. The performance of a part is an indication whether a part is fit for purpose.

In the present disclosure the wording, ‘part to be manufactured’ is used. The purpose of this wording is to refer, without any limitation, to an object that is produced for a specific function. The wording ‘part’ is typically intended to include alternative wording like ‘structure’, ‘article’ or ‘product’.

The method and system according to the present disclosure provide an integrated solution for improved part design and manufacturing. In the process of developing a part, the following steps can be identified:

- the design of the part,

- the selection of an adapted material, or combination of materials for the part,

- the manufacturing of the part,

- the performance assessment of the part and,

- the integration of the above mentioned steps. When using methods according to the prior art to develop a part, the overall process can be both time consuming and costly. One of the reasons for these lengthy and costly processes is the fact that the different steps of the process identified above are mutually related. For instance, the design of the part will have an effect on the material to be selected; the actual manufacturing process selected and used will have an effect on the performance of the part, etc. When during the development process, for instance, the design is amended to increase the performance of the part, this design amendment will automatically have an effect on other steps of the process, like on the selection of the most adapted material to be used.

Because of the mutual influence of the different steps of the procedure, in practice, the process for designing and manufacturing a part becomes very much an iterative process.

The disadvantages related to conventional product development cycles, include:

- long average lead-time from first design to prototype. In practice, this lead-time may be up to several months;

- the initial design will be very much designer experience biased, wherein the initial design is obtained with limited considerations from material and manufacturing constraints;

- the initial design will be obtained with limited or no knowledge of costing considerations, advanced materials and manufacturing constraints on the product;

- the development of the part will necessitate time consuming cross-team collaborations in an attempt to take into account elements such as design concept generation, modelling and optimization, risk assessment, performance assessment and additive manufacturing process simulation;

- time consuming and cumbersome design iterations to meet all the design and manufacturing requirements; and

- time consuming physics-based analysis methods requiring high resources, such as computational hardware and skilled users, to obtain reliable results.

Fig. 1 schematically shows the workflow of a conventional development process 100 for a part. The workflow 100 shown in Fig. 1 is also referred to as the ‘’as designed” approach. The workflow 100 as shown in Fig. 1 is specifically adapted for additive manufacturing development of a part.

In a first step 101 an initial part design is obtained. In a second step 102, an initial performance evaluation is performed, combined with a possible design optimization. In a third step 103 a Design for Manufacturing (DFM) evaluation is completed. This DFM evaluation describes the process of designing or engineering a part in order to optimize the manufacturing process, for instance to reduce its manufacturing costs.

In a fourth step 104 the final design is obtained and a final ‘as designed’ performance evaluation of the part is performed.

In a fifth step 105 the design is finalised and in a sixth step (106) 3D print data relating to the finalised design are obtained.

In a seventh step 107 an initial part is manufactured by means of an additive manufacturing process.

In an eighth step 108 possible manufacturing issues are identified and resolved and, if required, the design is modified.

In a ninth step 109 the part is further inspected, in view of the assessment of the manufacturing readiness level (MRL).

Thereafter, optionally, the MRL approval is obtained in a tenth step 110. Thereafter, or directly after step 109 the process is completed with a part testing and qualification step 111.

The workflow 100, as described with reference to Fig.1 , has several shortcomings. For instance, the ‘as designed’ approach as presented in Fig. 1 does not allow for the inclusion of the ‘’as manufactured” part design nor the inclusion of manufacturing physics and process data in the part design.

Moreover, the workflow 100 does not include real-time predictive manufacturing quality assurance and it’s mapping into the part development.

A further shortcoming is the fact that the cyber security, of which the scope is represented by steps 105 and 106 of workflow 100, is restricted to design and 3D part print data. The cyber security of additive parts extends from CAD modelling to slicing-tool path generation.

However, this is not the same as applied manufacturing process quality control.

Additive manufacturing is especially vulnerable to cyberattack, due to the reliance on digital files and connectivity, and the impact on multiple parties through the supply chain. Designs are created digitally, and, via connected printers and production lines, can theoretically be manufactured anywhere, at any time, and by anyone with the means to do so. Additive manufacturing’s reliance on digital files and connectivity can also open the process up to entirely new types of cyber threats, from product malfunctions to intellectual property theft and brand risk.

Therefore, a further objective of the method and the system according to the present disclosure is to improve the cyber security of the workflow for developing a part. To improve the process for the development of a part, according to the present disclosure each of the key steps in the part development are integrated. That means that the design step, the step of material selection, the step of manufacturing and the step of performance assessment are all integrated in one development platform. The platform is based on data obtained by bi-directional communication during all of the mentioned development stages.

This means that with user inputs on part requirements an optimum solution for design and manufacturing can be obtained by creating or modifying existing parametric CAD. This means that design intelligence can be used for future design optimisation due to the fact that historic data of earlier designs are collected and stored. Data relating to key parameters of materials can be used for optimisation of the material selection. Moreover, the same data can be used to identify optimised material properties and behaviour to thereby develop optimised materials. Data obtained during the manufacturing of the part can be collected and stored and used during the assessment of defaults. This can help to optimise the manufacturing configuration. Other effects of the integration of the development steps will be described in detail below.

The aim of the integrated approach of part development according to the present disclosure, backed by digital methods, is to propose a solution to overcome at least partially the above mentioned disadvantageous of existing part development methods.

Integration

Fig. 2 shows in a schematic manner the holistic approach for a digital integrated workflow for the development of a part. As shown in Fig.2 data related to, traditionally subsequent, steps in the development phase are interconnected and influence each other mutually.

The method and system according to the present disclosure are able to improve part development, because of the following, not limiting features.

During the initial stages of the part development, design files are pre-screened by automatically communicating with a design database to review designs that are stored in the design database to identify a nearest design and evaluate suitability of this nearest design to meet new determined requirements. to allow this step of pre-screening of designs, in an earlier phase a design database should be created comprising all available data of existing designs in a format, which accommodates the mentioned pre-screening.

To be able to use the design database for the above-mentioned purposes, the design database will have the form of a trained design database, comprising design features, based on focused product families to iteratively create a design from scratch which fulfils determined new requirements.

A further functionality of the method and system according to the disclosure is to perform cost estimation and initial performance analysis on the generated design iterations to compare and choose the best suitable option for the design.

A further functionality relates to automatic communication and evaluation of different material options to select the best candidate for the chosen manufacturing option. In order to allow this functionality the method and system of the present disclosure comprises a database comprising key parameters for a number of materials. These key parameters can comprise parameters, which have been obtained by means of testing and/or by means of calculations.

A further functionality is the possibility to perform automatic manufacturing assessment using process simulation using surrogate model in order to reduce the execution time of the analysis process.

To take full profit from the method and system of the present disclosure, the above-mentioned design step, material selection and manufacturing process optimization should be approached as a multi-objective problem to obtain the optimum result.

Part of the method and the system of the present disclosure can be obtaining a virtual system model, or digital twin, of the part to be manufactured to capture and update design performance based on in-situ data from physical part printing. Moreover, reverse learning is possible in combination with design updates, based on IOT (Internet of Things) data, which are retrieved from existing qualified designs which are in service. Hence, with minimal manual intervention from a design engineer the method and system of the present disclosure facilitates the complete development cycle of a part.

Figs. 3 - 6 show in more detail different integration steps which are used to allow the improvements of the part development as described above. Fig. 3 represent an integration step 1 wherein as input part functional requirements are used, such as pressure load and life. Moreover, information about the material selection is used as input. A further input relates to manufacturing process constraints and data stored in a designs database.

The output of the integration step 1 is a number of possible designs and a comparison of these designs based on functional performance, production costs and ease of manufacturing. This outcome is further processed to obtain a selection of, for instance, three designs based on defined constraints and/or criteria. Moreover a geometry clean-up is obtained for neutral or CAD native file format export.

Fig. 4 represents an integration step 2. The input for the integration step 2 are the most promising designs, for instance, in native format. A further input is material information. In addition manufacturing machine information is used as input. A final input has the form of physics. The output of integration step 2 is a manufacturing ‘as designed’ performance comparison. A further output are trained surrogate models, prepared print job input files, postprocessing plans (HT,HIR, etc.) and a virtual system model. These outputs are processed to obtain a final design, based on production costs, manufacturing comparison and ‘as designed’ performance.

Fig. 5 represents integration step 3 wherein the input comprises: the final design iteration, a prepared print job file, design quality criteria and the Digital Twin for the design.

The output of integration step 3 is ‘as manufactured’ part quality, ‘as manufactured’ virtual system model, an updated virtual system model with part performance based on defects and process quality interaction data for the material database in the form of feedback for the materials database.

These outputs are processed to obtain an initial design with optimum material use, taking into account the required performance of the material at different areas of the part.

Fig. 6 represent integration step 4. The inputs for this integration step are a virtual system model, a physical print design, a qualification plan as per standard and reliability requirements.

The output of this integration step 4 comprise an updated virtual system model with actual part performance based on service loads, design feedback based on in-filed service and an as-maintained virtual design model.

These outcomes are processed to obtain design data that can be fed back to the design database. In order to build the system and to operate the method according to the present disclosure a large amount of data needs to be collected. These data concern, in particular, historic data concerning existing designs, data relating to materials to be used to manufacture a part and data relating to selected manufacturing processes.

Manufacturing

Fig. 7 shows a workflow 200 relating to a method and system for integrated manufacturing intelligence according to the disclosure, which is adapted for robust additive part design and manufacturing. Fig. 7 specifically shows that data relating to the manufacturing of a part can be collected, stored and used for improving the part development process.

The workflow 200 allows for integration of additive manufacturing (AM) build design and process modelling tools into part performance prediction tools. This will enable back and forth transfer of simulation outputs from part performance and process modelling tools.

The effect of these measures is that an “as manufactured” part geometry can be developed from simulation and manufacturing data. The objective of this is to use this "as manufactured" part geometry for subsequent design phases as well as for future manufacturing records of parts obtained by means of additive manufacturing. This means that the design phase of a part can be improved and that the knowledge of part design can be stored in a database to allow the use of this knowledge for future part design. Process modelling tools are equipped with analytical failure models for high fidelity manufacturing process modelling.

The workflow 200 also allows for the development of Digital Twins, Reduced Order Models and Surrogate model tools for mapping manufacturing process physics as intelligence into the part design phase through the manufacturing process.

Moreover, the workflow 200 allows for the application of an integrated real time additive part quality assurance module. This module is developed based on Machine Learning, Artificial Intelligence (Al) and image processing technology. This quality assurance tool is applied for diagnosis and predictive monitoring of manufacturing process and control.

The workflow 200 is used to create a central knowledge database comprising process signatures for each printed part. This means that data about the manufacturing process of each printed part is generated during manufacturing and stored for future use. These data comprise, for instance, information about heat transfer during the printing process. This data can be used, for instance, for future failure investigations for each of the printed parts.

The cyber security solutions for the workflow 200 encompass quality assurance modules to address possible threats during manufacturing / processing operations.

The workflow 200 of Fig. 7 starts with an initial part design represented by step 201. In a subsequent step 202, part performance modelling is performed. After step 202 the data obtained in step 202 can be used, if needed, to improve the initial part design of step 201 . Alternatively, after step 202 a step of part process modelling 203 is performed. This means that data relating to the actual manufacturing process are added to further improve the part design.

As shown in Fig. 7, step 203 of part process modelling is performed using also information taken from block 204, which represents a database comprising data concerning process optimization.

As shown in Fig. 7, the information obtained during step 203 and the data stored in 204 can be used to complete step 202 relating to the part performance modelling.

After the completion of step 203, a final design is obtained in step 205. In this step, 3D part print data are developed, needed for the actual manufacturing of the part.

Block 206 schematically represents the additive manufacturing set up. Reference a refers to powder bed monitoring and control solutions. Reference b refers to melt pool monitoring and control solutions. Reference c refers to a heat source, such as a heat source laser or an electron beam. Reference d refers to the machine bed on which the part is being printed.

As shown in Fig. 7, the actual additive manufacturing device is directly connected to an Integrated Quality Assurance Module 207. As indicated by means of the double arrow in Fig. 7, information from the Integrated Quality Assurance Module 207 can be forwarded to the additive manufacturing device 206. At the same time, data obtained during the manufacturing of a part can be forwarded to the Integrated Quality Assurance Module 207. The collected signal data from machine is processed where in defect diagnosis is performed. Collected data is fed back to design module.

The Integrated Quality Assurance Module 207 is connected, in Fig. 7, to both step 201 relating to the initial part design and to block 204, relating to the process optimization.

This means that the data stored in the Integrated Quality Assurance Module 207 can be used to improve the design process in step 201 for obtaining an initial part design. This means that the manufacturing data stored in the Integrated Quality Assurance Module 207, allow historic manufacturing data to be used to improve the initial part design.

Moreover, the data from the Integrated Quality Assurance Module 207 can also be used in block 204, relating to the process optimization. In other words, the workflow 200 allows for part design and manufacturing with constant learning, wherein data relating to earlier part design and manufacturing are used to improve subsequent part design and manufacturing.

The Integrated Quality Assurance Module 207 is further connected to block 208, representing part inspection and MRL approval. Block 208 itself is connected to block 209 relating to the manufacturing planning and block 210 relating to the part testing and qualification.

According to the disclosure, quality assurance of the selected part during identified manufacturing process is maintained through intillegence diagnosis of selected manufacturing process parameters and their influence on part quality. This also includes analytics of manufacturing data for predicting part performance through connection of process data with physical properties and performance.

Fig. 8 shows part of the workflow 200 and shows in more detail the interaction between the additive manufacturing device shown in block 206 and the Integrated Quality Assurance Module 207. During the manufacturing of a part, data relevant for the manufacturing process, such as powder bed images and melt-pool data are forwarded to the Integrated Quality Assurance Module 207. The Integrated Quality Assurance Module 207 will comprise ML/DT/lmage processing based modules, "as manufactured” 3D part profiles and process maps connecting process parameters and defects.

In use, the Integrated Quality Assurance Module 207 provides process monitoring and control data to the additive manufacturing device shown in block 206. The information from the Integrated Quality Assurance Module 207 will be used to obtain an ‘’as manufactured” 3D part profile and CAD model. This is shown in block 211 . Moreover, the Integrated Quality Assurance Module 207 is used for obtaining experimental information based process maps for process optimization as shown in block 212.

As shown in Fig. 8, the combined information from blocks 211 and 213 is used to integrate with the part design platform 213.

The main advantages of the workflow 200 as shown in Figs. 7 and 8 is that it provides an integrated platform for part design with “as manufactured” part performance estimation. The workflow further provides part and support failure criteria for simulation of additive modelling process modelling.

The workflow also allows mapping of the “as manufactured” 3D part model, provide process maps developed from powder bed level, melt pool level monitoring assets and the additive manufacturing device inputs.

The image processing, which forms the first instance of the data collection mechanism, can be connected with a workflow in situ.

Quality assurance

Workflow 200 allows the application of integrated cloud based quality assurance module leveraging, image processing and will lead to reduced order models for melt pool and powder bed level data.

Workflow 200 also leads to a more complete cyber security solution, able to protect the part design and manufacturing process. Further it is possible to envision cloud based solutions enabling personalized threat alerts for individual printed parts.

The technical effect of the above-mentioned advantages is a significantly reduced part development cycle time and cost due to integrated design and manufacturing insights. Further, there will be an improved manufacturing quality, FPY and productivity due to more intelligent manufacturing.

A further effect is the possibility to optimize the manufacturing resources, reduce manufacturing waste and/or reduced CONC costs.

Finally, there is a reduced risk of cyber threats due to data protection and threat detection in the design and manufacturing process.

Design

Fig. 9 shows a workflow 300 wherein a step of pre-screening of existing designs can be used to improve and accelerate the part development. Fig. 9 refers, as an example, to the development of a jet pump. The workflow 300, comprises a currently used workflow 301 for the development of a part. The workflow 301 consists over several steps and requires time consuming communication between a part designer and several other experts, each responsible for various parts of the analysis.

Workflow 301 starts with step 310, which refers to CAD modelling. When receiving a new requirement for a new part, iterations of the design are needed to review whether the design geometry will satisfy all the performance and manufacturing criteria. During this process the CAD geometry is modified and checked for different performance metrics related to flow, structural, thermal and electrical performance, etc. in CFD analysis and then structurally analysed to determine compliance with requirements.

Once this is done, the CAD geometry might be topologically optimized, if needed. The result may be analysed again with a performance assessment tool, as shown in steps 311 and 312. Once a satisfactory design is produced manufacturing process simulation is performed in step 313 which might require using more than one simulation tool to define the orientation of printing, the required support structure and perform the layer by layer additive manufacturing simulation. Depending on the results the design could be modified again to improve on the build step.

As shown in Fig. 9, the development of the part is improved by adding a workflow 302 which relates to a phase of pre-screening of existing designs. This means that after determining new requirements for a part, indicated by step 303 in Fig. 9, the part and the new requirements are used as input for an automatic search in a database comprising the historic designs of parts.

In order to allow this automatic search, firstly a database should be created comprising data relating to designs should be collected in order to create a database. This database should be completed by adding the data relating to each new final design that is obtained.

In the example of the development of a part, after receiving new requirements for the jet pump in step 303 these new requirements can be used to review which of the design in the database comprises data which could be of relevance for the new jet pump to be developed. The search in the database is executed using intelligent algorithms to compare, for instance, data relating to cavitation, pressure ratio, efficiency, etc. in CFD analysis. Thereafter, the design can be structurally analysed to determine compliance with the determined requirements. The outcome of the above mentioned comparison can be that a direct match is found, which means that the design for the new part and for the new requirements is readily available in the design database. A second result can be that there is a partial result, which means that existing design provide partial solutions for the new requirements and that further development of the existing design is need to obtain an acceptable final design. A third result may be that the design database does not comprise any matches and that methods are deployed to create a clean slate design.

It is clear that the workflow 301 can be executed more efficiently, after the above-mentioned pre-screening phase 302.

Material selection and development

An important element of the above-described integration of different steps of the development of a part, is the use of material intelligence. The data available relating to different materials are used to facilitate the material selection process.

The essentials of material intelligence are described with reference to Fig. 10. Fig. 10 shows that material intelligence comprises continuously improving and completing a database comprising data relating to materials by using artificial intelligence and platform integration.

Reference 41 in Fig. 10 refers to a materials database. This materials database 41 is used during the development of a part, to allow efficient and adequate material selection for manufacturing a part. To allow this, the data in the materials database are consolidated in a structured format. The data are visualized in a manner to allow review of the most useful and relevant information. The aim of the materials database 41 is to provide a one stop location for designers, analysts and researchers to represented to identify materials and their respective properties. The use of the materials database can include calibrating the initial computational material model is with a set of material compositions. This validated computational material model is used for generating key performance parameters for a defined range of material compositions.

As represented by reference number 42, the data in the materials database 41 is used for intelligent screening and selection by means of Artificial Intelligence (Al). Al can also be used for material optimization and for real time property prediction, taking into account the actual manufacturing process is used to process the material.

Reference 43 refers to platform integration. That means that there is a seamless information exchange with third party tool relating to design, performance assessment, manufacturing and quality. The platform integration also allows for compatibility assessment and for providing scalability. The material intelligence as schematically represented in Fig. 10 allows material selection and material development.

The material selection includes identifying the design needs and based thereon enhancing material screening and selection. Moreover, the material selection takes into account performance assessment and manufacturing variability. Moreover, the material intelligence leads to material optimisation using analytical and digital characterization techniques.

The above-mentioned material development refers to building material recipes. It also refers to deep learning techniques to optimize material ingredients and compounds. Further, it refers to the possibility to generate meta data adapted for new material development.

Fig. 11 represents a workflow 50 relating to material selection and a related workflow 60 relating to material qualification. Fig. 11 shows the interaction between the product qualification, the manufacturing intelligence, the process assessment, design intelligence, material intelligence and the interaction with a user.

The materials database 51 is at the centre of the material selection workflow 50. A user can query the material database 51 using basic material needs. The database 51 provides basic material properties to iterative process 52, related to the design. This iterative process 52 receives input from the material database 51 and provides design feedback to the same database 52.

The materials database 51 further provides information to the iterative process 53 relating to the manufacturing. This iterative process 53 also receive information from the iterative process 52 relating to the design. The iterative process 53 provides process feedback to the materials database 51 .

The material database 51 is connected to a database 54, which relates to Digital Material Qualification (DMQ). The process of material qualification is a process that normally takes place between a material supplier and the customer. The objective is to make sure that the materials meet the requirements set by the customer. Material qualification is a useful tool to accelerate the introduction of new materials and is a process that ensures that new materials can be considered during future part development.

For the purpose of completeness, it is noted that material qualification differs from material development and material verification. Material development is the development of a new material not in existence before. This may be triggered by product or market requirements. The development process is followed by the qualification process. Material verification, on the other hand, is the process of making sure that the material used is the right one for a specific purpose.

Database 54 is connected to iterative process 55 that is related to the performance assessment. The iterative process 55 provides feedback to the DMQ database. As shown in figure 11 , the iterative process 55 also receives input from iterative process 52 relating to the design.

Workflow 60 relates to material qualification. The DMQ database 54 is at the centre of workflow 60. The input from a user is provided to database 54 in the form of next generation part requirements and/or material needs. The DMQ database 54 is connected to materials database 51 and to a Artificial Intelligence (Al) and Machine Learning (ML) module. The database 54 provides information to the performance assessment, as described above. The AI/ML module receives input from the DMQ database 54 and from the part qualification and the manufacturing process feedback.

Fig. 12 shows workflow 70 relating to material development, which is enabled due to the above-described integration of then part development process.

Box 71 in Fig. 12 comprises a first materials database 72 and a second material database 73 and the DMQ database 74. The first materials database 72 comprises material data obtained via computational processing. The second materials database 73 comprises material data obtained via experimentation. The data in box 71 are completed with data received from manufacturing processes, indicated in box 75.

The input received for box 71 is a determined amount of data relating to new material requirements.

The data in box 71 are used for disciplined experimentation, as shown in box 76. Box 76 receives input from material constituents, box 77, and from qualification inputs, box 78.

The data obtained in box 76 are used to obtain unified charts, box 79, which permit feature extraction. The extracted features are fed into box 80 which relates to model development and verification and which comprises regression, clustering and classification. Data taken from box 81 are used for experimental validation. If the validation is successful the search for a new material is completed. If the validation is not successful, the data from box 81 are fed into box 79, as shown in Fig. 12.

The data relating to each new material are fed into box 82, which relates to product performance evaluation. The data relating to performance feedback are fed into box 83, which relates to material screening and selections. The outcome of box 83 comprises constituents refinements, which are fed into box 77.

The effect of the above-mentioned features is that a single integrated platform is obtained with seamless material information exchange across different technology domains. The consolidated database, box 71 in Fig. 12, comprises both experimental and digital material properties.

The described method and system allows for two way interaction enabling design, manufacturing combinations for material screening and selection. Moreover it allows for real time property prediction to meet product qualification needs. In order words, the method and system according to the disclosure provides an in-built intelligence for material optimization.

Performance

Part of the performance analysis is the use of an custom developed application. The function of the application is to integrate all structural load cases, to thereafter identify a maximum stress envelop, divide the volume into mutually exclusive stress zones and select optimum parameters for the manufacturing process for the volume and export a geometry file to a 3D printer.

In the above mentioned process for a new part a design is selected and for the design the following steps are executed:

1) a computer model of the part and data relating to different functional load cases for the part are processed in a computer to obtain result files;

2) the results of the different result files are combined; this includes the step of mapping all the results into one mesh in case of multiple analysis files having a different mesh;

3) in a subsequent step the maximum stress envelope is calculated for the different load cases;

4) based on the maximum stress envelop critical zones are identified and the volume of the part is divided into mutually exclusive zones; 5) the final step comprises the selection of the print parameters for a 3D printer. This includes identifying optimum process parameters for each critical zone and exporting the geometry file for printing and process parameter information. According to further examples, wherein the selected manufacturing process is an additive manufacturing (AM) process, the method further comprises the use of a powder melting model to predict melt-pool attributes. These melt-pool attributes include the width, depth, shape and length of the melt-pool. The powder melting model is also used for transient temperature profile depicting temperature gradient and cooling rates. It is possible to achive a level of accuracy of more than 85%.

Additionally, solidification and solid-state transformation modelling can be used to predict solidification microstructure morphology, such as micro-segregation, dendrite size, dendrite shape, dendrite arm spacing, phases, grain type and grain size. These parameters can be predicted for different laser heat energy input levels. In practise, the microstructure morphology, such as phases, phase volume fractions, grain shape and size, can be predicted at different heat treatment conditions with an accuracy of more than 90%.

Further, homogenization modelling can be used to predict the mechanical properties at as- built and heat treatment conditions. This prediction can be done with an accuracy of more than 85%.