Reliability is rarely thought about when things are going well and machines are running smoothly. But that can all change in an instant. In effect, once a machine is delivered and installed, reliability is perhaps the machine’s most important characteristic as far as the customers are concerned.
Poor reliability affects companies by adding downtime, maintenance costs and other downstream effects. This is easily understood by anyone who’s experienced a faulty piece of equipment. It’s not just fixing the machine, which is enough of a burden; there are always additional and costly consequences to deal with.
An unreliable machine destroys the credibility of the company that made it and can easily led to lost customers and sales. In competitive marketplaces, price and performance are important and can differentiate one company from the rest of the field. But a reputation for designing and building reliable machines is another way to set a company apart from its competitors. Many companies have strived to become synonymous with reliability.
Becoming known as a company that supplies reliable equipment cannot be done by using clever marketing. For maximum effect, the drive for reliability must permeate the company.
Of course, one critical area is design.
Designing for Reliability
Design is the most important factor in a machine’s reliability. Engineers often ignore reliability until too late in the design cycle. For example, many engineers only consider reliability late in the validity portions of the design cycle rather than during the concept and early design phases.
Once the concept has been agreed to, it becomes more costly and inefficient to add reliability measures. But it is often less expensive to design for reliability than it is to test for reliability.
There are at least two proven and systematic methods for recognizing possible reliability issues during the design process: Reliability Block Diagrams (RBD) and Failure Modes and Effect Analysis (FMEA).
An RBD lays out a model of the machine, listing reliability at each component. Engineers must be sure to follow the proper reliability path, which may differ from the control path. For example, an RBD for a car’s drivetrain might look like the block diagram below.
At each block, the reliability of the individual component is determined. And they feed into an overall reliability number.
An RBD is simple to understand and can quickly uncover potential reliability issues, as it easily exposes “weak links in the chain.” But it can also be too simplistic for some machines, as it does not consider relationships between components. Does reliability of any of the blocks depend on how they are configured in a certain path?
FMEA systematically identifies each failure mode of a machine or process. Examining failure modes in detail can also uncover other shortcomings in the design. This includes the underlying failure mechanism as well as ways to eliminate it or reduce its chances. (Risk Priority Number, for example, is determined by multiplying the severity, occurrence and detection factors, as seen below. The resulting RPN gives designers an idea of how much of a problem the failure mode will be.)
Addressing severity, occurrence and detection in the design phase is critical for designing reliable devices. If the RPN is high, engineers have two options: eliminate the failure mode or change one or more of the factors to get a lower RPN.
The best course of action is not always clear. Sometimes all that are needed are a few small design tweaks, sometimes engineers can add an additional control mechanism and sometimes the design team must go back to the drawing board.
FMEA is generally a thorough exploration of all a machine’s failure sources. Once it has been conducted, results can be used for replication on similar machines. Better understanding of the failure modes better can greatly help in current and future design. FMEA results help maintenance technicians understand when a machine is breaking down. This lets them respond faster and more accurately, and ultimately improve reliability.
Unfortunately, FMEA examines every possible failure mode, so it can be tedious, time-consuming and expensive. FMEA effectiveness is also highly dependent on the expertise of the people performing the analysis. Therefore, it takes people with a high level of experience to execute it.
Improving Reliability
Once the design team use RBD, FMEA or some other form of analysis to get a firm grasp on a design’s reliability weaknesses, it can more effectively address the reliability concerns. Common methods for improving reliability include applying Reliability Centered Maintenance (RCM) and focusing on proactive maintenance methods such as Condition Based Maintenance (CBM) and Predictive Maintenance.
RCM is like FMEA, but it goes further. It takes the failure modes from FMEA and develops maintenance strategies to address failures. RCM leads the team through each failure mode, where it determines the best maintenance strategy to prevent the failure. Most commonly, RCM is done after equipment is operating. Performing at the design stage, however, could lead to solid insights into reliability improvements.
Like FMEA, RCM is a systematic approach to treating failure modes through prevention. For example, if the designers, know a clogged filter will reduce air flow and damage an engine, the RCM response might be to schedule a filter replacement every three months. Learnings from one RCM program can also be used elsewhere.
But successful RCM require resources, training and dedication. A company should be sure it can fully support the strategy before it embarks on it. And like FMEA, it requires some expertise to develop RCM.
NASA once used RCM at its Marshall Flight Center and maintenance costs were decreased, existing equipment life was improved and energy costs were lowered, resulting in a savings of over $300,000. If these savings can be made after implementation, then using RCM in the design phase can surely yield benefits. If the design team works through FMEA for a new machine, the next logical step is RCM.
CBM uses real time machine conditions to determine when maintenance is needed. This is done by putting temperature, vibration or some other type of sensor on relevant areas of the machine and tying them into control loops or external databases. Naturally, this approach can be taken in the design phase. Even though it adds a relatively small amount of cost to the product, it would give end-users much better predictors of performance and reliability.
CBM tracks data not always discernable by human senses. It can remotely monitor equipment while it is running, saving time and reducing disruptions. But CBM is more costly for customers and requires more upfront setup and configuration. And here will be a learning curve in which the company establishes sensor thresholds. Training is also required. When does the maintenance team need to act? This is not easily understood.
CBM, when properly executed, reduces breakdowns and regular maintenance. One source puts CBM savings at 12% in the first year, with a reduction in failures exceeding 25%, and a 94% improvement in machine availability.
A simple example of CBM is adding vibration sensors to motors. By tracking the vibration frequency and setting up an alert at an appropriate level for action, you can react quickly to adverse conditions and prolong the life of the motor.
As many have stated, maintenance, repair and operations deserve a higher priority than they usually receive, especially during the design phase. If reliability is considered early in the design process, equipment will surely be better off in the long run. Reliability can become strength of the design and a method to differentiate a company’s products.
Bryan Christiansen is the founder and CEO of Limble CMMS.