Specifically, by performing FMEA, weaknesses in a product can be identified before the product design or manufacturing process is actually realized, and defects can be identified at the prototype stage or before high-volume
production.
FMEA was first developed by NASA as a set of analytical models, FMEA is a practical problem-solving methodology that can be applied to many engineering
fields, and is now being used by many of the world's automobile manufacturers and electronic manufacturing service providers (EMS) for the management and monitoring of design and production processes.
Introduction to FMEA
There are three types of FMEA, system FMEA, design FMEA and process FMEA, and in this article, we will focus on process FMEA.
1) Determine the technology that needs to be involved in the product, and the problems that can arise, including all of the following aspects:
New systems, products, and processes to be designed;
A review of existing designs;
A review of existing designs;
A review of existing designs. p>
Improvements to existing designs and processes;
Retained use of previous designs and processes in new applications or new environments;
Formation of the FMEA team.
The ideal FMEA team should include representatives from all interested parties, including design, production, assembly, quality control, reliability, service, purchasing, testing, and suppliers.
2) Document the serial number, date, and changes to the FMEA, keeping the FMEA always a real-time field record that changes according to the actual situation,
It is important to emphasize that the FMEA document must include the date it was created and updated.
3) Create a process flow diagram.
Process flow diagrams should be developed in accordance with the sequence of events and the requirements of the technical process. Process flow diagrams are required for the implementation of an FMEA, and they should not
be easily changed in general.
4) List all possible failure modes, effects and causes, and process controls for each operation:
4.1 For each process in the process flow, the possible failure modes should be identified.
For example, in the case of the surface mount process (SMT), issues may include, based on engineering experience, solder ball control, solder paste control, the type of soldermask to be used, and the design of the pad pattern for the component.
4.2 For each failure mode, one or more possible failure effects should be listed.
For example, soldermasks may have to affect the long-term reliability of the product, and therefore should be noted in terms of possible effects.
4.3 For each failure mode, one or more possible causes of failure should be listed.
For example, possible factors affecting solder balls include pad pattern design, excessive solder paste humidity, and solder paste volume control.
4.4 Existing process controls are based on currently used methods of detecting failure modes to avoid some of the root causes.
For example, existing process controls for solder balls may be automated optical inspection (AOI), or a well documented control process for solder paste.
5) Ranking the frequency, severity and detection level of events:
5.1 Severity is an assessment of the impact of possible failure modes on the product, with 10 being the most severe and 1 having no impact;
Frequency of events is to be documented in terms of how often and how often a particular cause and mechanism of failure occurs.
A value of 10 indicates that it is almost certainly going to happen, with a process capability of 0.33 or a ppm greater than 10,000.
5.2 The detection level evaluates the chances of detecting the failure mode for the proposed process control, with a value of 10 indicating that it cannot be detected, and 1 indicating that it has been detected as a defect by the current process control
.
5.3 Calculate the riskprioritynumber (RPN).
The RPN is the product of the frequency of the event, the severity and the level of detection, and is used to measure possible process defects so that possible preventive measures can be taken
to reduce critical process variations and make the process more reliable. Corrections to the process should first be focused on those areas of greatest concern and risk.
With an RPN of 1,000 at worst and 1 at best, the best way to determine where to start is to utilize a pareto chart of the RPN and screen for items that have a cumulative
rating well below 80%.
Recommend the responsible programs along with completion dates, and the ultimate goal of these recommended programs is to lower the grade by one or more. Rescue programs are considered from time to time for serious problems, such as:
a product with a failure mode impact with a risk level of 9 or 10;
a product with a failure mode/cause event with a high level of occurrence and severity;
a product with a high RPN, etc.
After all the salvage measures have been identified and implemented, a stabilization period is allowed, and then the frequency, severity, and detection
level of the revised events should also be reconsidered and ranked.
FMEA application
FMEA actually implies behavior before the event, not after the fact remediation.
Therefore for best results, it should be done before the process failure mode appears in the product. The five phases of product development include:
Planning and definition, design and development, process design, pre-production, and high-volume production.
As a major EMS provider, Flextronics International already utilizes FMEA management in production process planning and control, and introducing FMEA management early in the product
is critical to producing a high quality product and documenting and continuously improving the process. For most of the company's customers, FMEA management is used when the product is moved to the production center after the design and
production process have been fully defined.
Handheld Product FMEA Analysis Example
After this new product introduction (NPI) launch, an FMEA team is formed, consisting of a production director, process engineer, product engineer, tester, quality engineer, material buyer, and project manager, with the quality engineer leading the team. The goal is to strengthen the quality control points in the initial Manufacturing Process Instruction (MPI) and Test Process Instruction (TPI) while the team
gains a better understanding of the product. The main tasks during and after the first meeting generally include:
1. The process and production engineers present a step-by-step process flow diagram, where the function and requirements of each step of the process need to be defined.
2. Together, the team discusses and lists all possible failure modes, all possible effects, all possible causes, and current process controls for each step, and ranks these
factors by RPN. For example, for all possible failure modes of missing solder paste in a screen print operation, the existing process
controls are Stencil Design, periodic cleaning of the stencils, Visual Inspection, Preventive Maintenance (PM) of the equipment, and Preventive Maintenance (PM) of the equipment. Preventive Maintenance) and solder paste viscosity checks. The process engineer includes all current control points in the initial MPI, such as stencil design studies
research, determining stencil cleaning, frequency of visual inspections, and solder paste control.
3. The FMEA team needs to perform a targeted audit of the existing production line against the control nodes in the MEA document, synthesizing the current line setup and other issues
. For example, the location of the drying box, the audit team recommended that it should be placed near the Fine-pitch Placementmachine to facilitate
handling of humidity-sensitive components.
4. FMEA follow-up activities in the completion of the general structure of the NPI, can be carried out FMEA follow-up meeting. The meeting includes a synthesis of existing process controls and quality reports from the NPI's broad
consistent structure. The FMEA team re-ranks the RPNs, with the top three major defects considered first for each step, and determines the recommended
recommended program, responsibilities, and target completion dates.
For the surface mount process, the top two defects are solder ball defects and tombstone defects, and the following solutions can be recommended to process engineers:
For solder ball defects, check the stencildesign, check the reflow profile, and reflow preventive maintenance (PM) records.
Check screen printing accuracy and pick-and-place machine placement accuracy.
For tombstone defects, check screen printing accuracy and pick-and-place machine placement accuracy;
Check direction of return flow; study possibility of terminal contamination.
The process engineer's report indicated that a rapid rise in reflow temperature was the main cause of solder ball defects, and that contamination of the termination was the likely cause of tombstone defects, so a design of experiments (DOE) was set up for the next design validation test structure, and the design of experiments showed that
A supplier's components were showing signs of defects. p>a supplier's components have a higher likelihood of tombstone (tombstone) defects, so a corrective request for further investigation is issued to the supplier.
5. Any changes made to the product design, application, environmental materials, and production and assembly processes must be updated in the appropriate FMEA document.
The FMEA update meeting is a routine activity before the product goes into mass production.
Managing the FMEA during the mass production phase
As a historical document for process improvement, the FMEA is transferred to the production site in preparation for product release.
The primary role of the FMEA in the production phase is to check the FMEA document to get a handle on each control node prior to mass production, as well as to review the validity of the production line
All items that were not challenged in the NPI FMEA phase are naturally retained at the mass production site.
Pick-and-place machine accuracy is a major consideration after the process review, and the equipment department must validate the Cp/Cpk of the layout machine,
along with training to deal with misprinted boards.The FMEA team needs to closely monitor the first trial production, and quality validation of the production line should be done at the same time.
After the pilot production, the FMEA needs to hold a meeting to verify the existing quality control with the quality report of the pilot production, addressing the first three issues of each process.
The FMEA manages to document a continuous effort to improve the process and continuity, and the FMEA document should always reflect the latest state of the design, including any changes made after the production
process has begun.
Conclusion
Using the FMEA management model to identify risks in a project at an early stage can help electronic equipment manufacturers increase production capacity and efficiency, and shorten the time to market for their products.
Additionally, this model also allows various experts to examine the production process from all angles, which can lead to improvements in the production process.
The recommended solutions should be correctly corrected and produce considerable benefits. Changes to the process and design are required to avoid defects. The production process is studied using statistical
methods and constant feedback is given to the right people to ensure that the process is constantly improved and defects are avoided.