Search

EP-3663874-B1 - METHOD AND SYSTEM FOR OPTIMIZING A MANUFACTURING PROCESS BASED ON A SURROGATE MODEL OF A PART

EP3663874B1EP 3663874 B1EP3663874 B1EP 3663874B1EP-3663874-B1

Inventors

  • RUGGIERO, ERIC JOHN
  • TALLMAN, James
  • SALAPAKKAM, PRADEEP

Dates

Publication Date
20260506
Application Date
20191203

Claims (12)

  1. A method for optimizing a manufacturing process of a new component, the method comprising: executing, by a system configured to drive the manufacturing process, a set of manufacturing functions, the executing including: setting (204) up manufacturing directives for making the new component, the directives include a set of manufacturing functions for making the new component, wherein the manufacturing functions are based on geometrical constraints or a physics-based model; updating (214) the set of manufacturing functions based on a surrogate model, wherein the surrogate model is provided via training based on one or more parameters selected from the group of parameters consisting of: data relating to manufacturing parameters, customer constraints, and performance data obtained from instances of the same component as they are used in the field, the surrogate model being generated from running one or more machine learning algorithms, and wherein the one or more machine learning algorithms are configured to predict the expected life of one or more subsystems of a component; and manufacturing (216) the new component according to the updated set of manufacturing functions, characterized by : generating (306) a forecasting model for each of the individual components including a score for each component; sorting (308) forecasting models according to their scores according to their anticipated longevity; and deploying (310) components according to the scores.
  2. The method of claim 1, wherein the surrogate model is associated with an engine component.
  3. The method of claim 1 or 2, wherein the surrogate model is associated with a component in a hot gas path of an engine or with a cold section of the engine.
  4. The method of any of claims 1 to 3, wherein the surrogate model is updated according to a machine learning model.
  5. A system for optimizing a manufacturing process of a new component, the system comprising: a processor; a memory including instructions that, when executed by the processor, cause the processor to perform the method of any preceding claim.
  6. The system of claim 5, wherein the surrogate model is associated with an engine component.
  7. The system of any of claims 5 to 6, wherein the surrogate model is associated with a subsystem of an engine.
  8. The system of any of claims 5 to 7, wherein the operations further include executing an analytics module including a machine learning module.
  9. The system of any of claims 5 to 8, wherein the set of manufacturing functions represent an optimized set of manufacturing functions relative to another of set of manufacturing functions that are not based on the surrogate model.
  10. The system of claim 8, or claim 9 when dependent on claim 8, wherein the analytics module is configured to update the surrogate model via training.
  11. The system of claim 8 or any claim dependent thereon, wherein the analytics module is configured to update the surrogate model based on a set of manufacturing parameters, laser settings, part rotation speeds, EDM burn rates, and a robustness metric.
  12. The system of claim 8 or any claim dependent thereon, wherein the analytics module is configured to update the surrogate model without using a physics-based model of the new part.

Description

TECHNICAL FIELD The present disclosure generally relates to manufacturing processes. More particularly, the present disclosure relates to a system and a method for optimizing a manufacturing process based on a surrogate model of a part. BACKGROUND The typical lifecycle earnings curve of a newly launched product may exhibit three distinct phases. The first phase, which is the earliest, typically includes early field experiences in which customers adopt and evaluate product, and the manufacturer of the product gathers data regarding commercial and technical aspects of the product's in-field use. In the first phase, the manufacturer's earnings are typically negative, as the cost of manufacturing and deploying the product may exceed the revenues generated from early-adopting customers. The second phase, which typically begins at the point where earnings become positive, includes a high growth phase where the product is adopted by many customers and earnings grow at a rapid pace. In the third phase, a plateau in the earnings is typically observed, which suggests that the product has reached its maximum market potential. From a manufacturer's perspective, one of the key goals of a typical manufacturing processes is to provide value to customers. As such, parts are typically designed to provide the maximum durability possible, and typical manufacturing processes are optimized accordingly. However, typical manufacturing processes lack the capability to favorably alter the earning's curve without compromising durability. In other words, typical manufacturing processes cannot favorably mitigate costs without compromising performance. US2018137219 A1 discloses systems and methods for predictive modelling of an industrial asset. WO 2016/179455 A1 relates to a data-feedback loop from product lifecycle into design and manufacturing. US 2017/124448 A1 relates to a concurrent uncertainty management system. SUMMARY The invention is defined by the appended claims. Claim 1 defines a method and claim 5 defines a system. In the following, apparatus and/or methods referred to as embodiments that nevertheless do not fall within the scope of the claims should be understood as examples useful for understanding the invention. BRIEF DESCRIPTION OF THE DRAWINGS Illustrative embodiments may take form in various components and arrangements of components. Illustrative embodiments are shown in the accompanying drawings, throughout which like reference numerals may indicate corresponding or similar parts in the various drawings. Furthermore, the drawings are only for purposes of illustrating the embodiments and are not to be construed as limiting the disclosure. Given the following enabling description of the drawings, the novel aspects of the present disclosure should become evident to a person of ordinary skill in the relevant art(s). FIG. 1 illustrates a process according to an embodiment.FIG. 2 illustrates a method according to an embodiment.FIG. 3 illustrates a method according to an embodiment.FIG. 4 illustrates a system according to an embodiment. DETAILED DESCRIPTION While the illustrative embodiments are described herein for particular applications, it should be understood that the present disclosure is not limited thereto. Those skilled in the art and with access to the teachings provided herein will recognize additional applications, modifications, and embodiments within the scope thereof and additional fields in which the present disclosure would be of significant utility. An embodiment provided can include a system that is specifically configured to perform manufacturing optimization. For example, and not by limitation, such a system may be configured to provide a surrogate model for driving an optimized manufacturing process. The surrogate model is generated from running one or more machine learning algorithms using inputs such as manufacturing parameters, environmental factors, customer de-rate, and route structures. The one or more machine learning algorithms then predict the expected life of one or more subsystems of a component. For example, and not by limitation, the one or more machine learning algorithms may provide a predictive model of a hot gas path component for a given customer and ESN. The surrogate model may take in the new manufacturing parameters, expected route structures, customer de-rate, and environmental factors and make a recommendations that mitigates a trade-off typically encountered in the manufacturing process. One such trade-off may be, without limitation, the manufacturing cost versus the expected durability of the component. In yet other embodiments, the surrogate model may be representative of a component in a cold section of an engine. For example, and not by limitation, such a component may be compressor blades, fan blades, or fuel nozzles. Generally, the surrogate model may be trained through algorithms based on machine learning and/or the transfer function between manufacturing parameters (e.