What tools and methods are effective for preventing engineering design failures?

To prevent engineering design failures, it's crucial to follow a robust engineering design process. This process should be systematic, iterative, and collaborative, involving all relevant stakeholders, including engineers, designers, clients, and users. It should also include clear and comprehensive requirements gathering, design reviews, testing, and validation. Here are some specific steps that engineers can take to prevent design failures: 1- Clearly define the problem and objectives. 2- Gather comprehensive requirements. 3- Conduct a thorough design review. 4- Perform thorough testing and validation. 5- Use risk management techniques.

One tool I found useful is the design checklist in form of questions If the questions are elaborated in written form by engineers with expertise in the field and if they are used at least 2 times in the course of a project, they become an effective way to make explicit the implicit engineering knowledge of a design group. The use of the checklist is very useful to train quickly the young colleagues to minimise mistakes due to their short time experience

The project commonly is complicated. They have many various input data that need to be proceed being requirement data. To reduce data error, we require share eng data for project multi member for their review then give feed back. Concern in making proper system for design input/output between project multi discipline, ideally some stage engineering data issuance for example data for information, review, approval and construction (final data)

We must also look at standards and best practices. For metallurgical look at the ASME and ASTM standards, they are there for a reason! For software look at the IEEE standards for coding and testing. Also for coding look at the GAMP standards for software development, documentation and testing, and do it! No short cuts!

Design failures can be caused by a) failure to articulate real, achievable, testable customer-ratified performance objectives during the requirements development process and; b) making cost savings a higher priority than the actual performance requirement itself such that it leads to errors in judgment by program decision-makers.

When it comes to design failures, the severity of the failure can vary. Some failures are minor and only lead to inconvenience or minor damage. However, other failures are much more serious and can result in significant damage or injury, such as a bridge collapse. It's important to note that multiple factors, not just one often cause design failures. A bridge may collapse due to poor design, inadequate construction, and overloading. Therefore, it's crucial to investigate all possible causes of failure to create effective prevention strategies. Engineering is a complex field, and there's always some level of uncertainty. However, by understanding the different types of failures, engineers can minimise the risk of failure.

Engineering design failures and quality or workmanship failures are in different categories. Companies are sometimes too quick to attribute failures to quality, workmanship or application, and too slow to consider shortcomings in design. Quality or workmanship failures will usually show up as zero-hour or low time failures. Application or user induced failures will be statistical outliers, and can be distinguished by comparing failure rates and probabilities across the entire fielded population. A quality or workmanship failure will likely be distinct for a range of manufacturing lots or date codes. Engineering design failure rates are more likely to be consistent across the entire population and customer base.

I think categorizing "failures" as functional, structural, operational and aesthetic is only as useful as our approach to finding solutions. To me, a more important type is failure(s) of our design/development process itself. At regular intervals, we should be asking the question "is our design process effective?". Many of the failures observed in the field can be traced back to the "how" of our process. These are systematic failures that can be hard to define and quantify. We may "try" to find a "root cause", but unless we also look at how we run our design process itself, we may not really solve the true problem.

I think this type of failure definition is highly generalised. Indeed, I believe direct definitions could lead us to better solutions. For instance, over/underestimated design failure can have consequences on the operation without having a general perception of failure. Taking an oversized HVAC system into account, this system perhaps works without failure but with not efficient and general definitions does not name it a failure while not only has it imposed higher capital expenditure but also maintenance and environmental costs. So I prefer more straightforward topics regarding modern concerns that take efficiency and environmental impacts into accounts. Since failure meaning has changed in recent years.

While there are many applications for the prevention of failure, there is a wonderful tool which is very effective in understanding different modes of failures and analysing its impact on the performance of the product. The tool is known as FMEA - Failure Mode and Effect Analysis. With the help of this tool we not only can determine the various failure modes but also get a fair idea of the causes of failure and what would be the change in the product performance. In this tool we assign a severity value, level of occurrence and ease of detection for a particular failure mode and based on the product of these values define a RPN (Risk Priority Number) which in turn help in determining and implenting a better risk mitigation plan.

During the execution of projects that have found that many of engineering failures result from not documenting lessons learnt from previous projects. It is important to indoctrinate and sensitise new engineers on these failures.

> FEA > DFMEA > Right Material selection as per application of product > Pre defined standards if applicable > Prototype making > Testing of product on actual conditions.

One of the keys to success is talking to, working with and investigating the equipment market. Remember the suppliers of a equipment are the leading edge of what you need to buy and install. Trade shows and running technical comparisons are essential. Also talk to your engineering firms, they have a huge experience base of what worked and what failed. Use the industry and get your head out into the market.

What differentiate designers who make few design failures and those who keep failing endlessly in the name of iterations (looking at the mirror :) ) is the intimate knowledge of constraints. There are broadly 3 categories of such constraints that designers need to clearly identify before attempting to solve a design problem. Also, each designers has to tread certain paths in their work experience to be able to identify the constraints. 1) Manufacturing & Assembly constraints: Spend time in the shop floor observing and talking to machine operators & assembly personnel 2) User constraints: Spend time observing current behavior of users and test different mockups with them for feedback 3) Project constraints: Put yourself in the CEO's shoes

Drawing from my own journey in engineering, I've found that proactive methods for prevention are paramount in averting design failures. Rigorous design review meetings with diverse teams and comprehensive Failure Modes and Effects Analysis (FMEA) have been my go-to strategies. These methods help pinpoint weaknesses, ensure early detection of potential issues, and guide corrective actions. Regular cross-disciplinary collaboration and adhering to industry standards, along with continual education and root cause analysis, have been instrumental. My experience has shown that investing in robust prevention methods not only safeguards against design failures but also promotes more efficient, reliable, and cost-effective engineering outcomes.

It is interesting that not once does he mention prototyping, environmental shake&bake, humidity, temperature, altitude testing, functional testing, EMI radiation/susceptibility tests, and things like maintainability and repairability to make sure you're selling a product that can be used, can be repaired, does its job the best AND can be recycled.

As they say, learning from our mistakes is useful; but I think learning from the mistakes of others is even better! In the world of medical devices, we can learn a lot from FDA warning letters, safety alerts, and data available in the TPLC database. We should actively look for such information and analyze our own customer experience data to detect early signals before they become an issue. This is the essence of an effective post-market surveillance operation.

Using well-established methods and processes is advantageous for minimizing engineering failures. However, based on my overall experience, a comprehensive approach that takes into account and monitors every stage, from the initial concept to product realization and usage outcomes throughout the product's lifecycle, necessitates a strong sense of responsibility and motivation. To achieve this, effective engineering practices and innovative advancements, a sustainable and well-organized engineering teams are needed.

An example I have seen while working in a GERC. A reciprocating compressor driven by an engine using gas as fuel is referred to as GERC. This engine had a turbocharger installed in it. Air filter, bellow and 15 meter long pipe for air were present in the upstream of turbocharger as per design in the exact same order. The problem with that design is that this bellow was not clamped properly. One day during the normal operation this bellow got sucked into the air line causing rotor damage in the turbocharger. Bellows for the remaining GERC were replaced with a metallic bellow strengthening their clamping part. Learning from this incident enables us to proactively do these design changes and eliminate such incidents from happening again.

Each companies has different lesson learnt and mistakes made in the past regardless in term of cost, time or resources. Most important in lesson learnt is to record keeping and review so that this can be documented for continuous improvement. Some of companies may implement this in their ISO and internal procedures to make sure lesson learnt are being recorded and made reference for future note.

Appreciate the invaluable lessons that can be gleaned from past failures. One of the most effective tools for preventing design failures is learning from these setbacks. By maintaining a comprehensive database of past failures and the lessons learned, we've been able to proactively identify and address potential issues in our designs. These records serve as a constant reminder of what can go wrong and guide us in making improvements, ensuring that each design iteration is more robust and less prone to failure. In essence, embracing failure as a source of wisdom has been a cornerstone of our preventive approach.

In my experience working with downhole pumps for oil and gas wells, most of engineering and design failures are attributed to the data considered at the time of design. The life span of the any machine (like pumps etc.) is greatly effected by the changing environments that it has to encounter if they are not drained to handle this. Hence an engineer must be very clear about two things at the time of engineering design: 1> What is the target runlife for the equipment? 2> Does he or she has enough visibility of the data for the changing conditions that equipment will encounter during that targeted run life at the time of design. These are solely my views based on my limited experience.