Test To Failure: Why You Need This Reliability Test

Let’s look into what test to failure is, the benefits it provides, and why you will find it a useful tool in your product reliability testing arsenal for almost every product type, especially electronics. This post will also provide the goals of TTF, 5 key reasons why you need to implement it into your testing processes, and some common testing methodologies you can harness.

Prefer listening to reading?

Listen to the audio here.

What is test to failure?

A lot of the products we rely on today need to be reliable and durable, be that a pacemaker, mobile phone, laptop, automobile, etc. Most electronics firms do test to failure these days for their hardware (and software) in order to investigate what the product’s limits are.

There are usually 2 goals for this kind of testing:

To assess and confirm the product’s limits – i.e. if a product needs to handle 100 volts, how do you know if it is able to handle it if you do not test it above and beyond this voltage?
To give yourself a margin of failure – this is your product’s reliability, and in this case, if the product can handle up to 120 volts even if it’s going to be used with 100 volts, then you are giving yourself a 20% margin of failure.

Testing the engineering of electronic products in this way is common to provide peace of mind that if there are changes in voltage, for example, the product will survive and not short out. Therefore, to simulate, say, a lightning strike, the product will be subjected to a brief burst of high voltage and the results will determine how well it can cope. By necessity, a number of product samples may be damaged or destroyed while doing these tests.

The product is then designed to be able to handle similar conditions outside of its normal usage to add a layer of ruggedness or reliability (your margin of failure), and that may involve higher-rated components being selected, for instance. (01:28)

5 reasons why test to failure matters

People rely on electronic products these days and we can’t afford to have them break down on us. Cars are an obvious example. If the electronics on your vehicle fail, you’re going to break down which will be inconvenient and may even be dangerous! The same can be said for your cellphone or laptop.

There are 5 reasons why test to failure is so important:

To find the product’s weaknesses in the design and then eliminate them until we have gained an adequate margin of reliability (this may be through different components, materials, etc)
To enhance product safety as you cannot afford to put consumers in danger with, say, a battery that may catch fire or explode if the product is under extreme conditions.
It helps with quality control because you can test the product within its operational limits which you will know thanks to test to failure. Then when testing is done in the factory, staff are not put in danger.
Cost reduction may be achieved because if you know your product’s limits, you can tweak the design to be ‘durable enough’ without being excessively rugged which usually translates into more costly. The latter may be required for a moon landing, but if you use components that are a little lower-rated and less costly, you may find that the product works fine and still reaches a reasonable lifetime for your consumers. On the other hand, you may also find areas where better materials or components need to be used. While these items may cost more to purchase, reducing the cost of product returns will be a large cost reduction overall for your business.
It spurs innovation because you may find new ways of doing things or new materials or components during testing that have not been used before, perhaps by combining different materials to see if they do a job well. Patent these and you could have a lucrative invention on your hands or, at least, you have a really great way to make your product more reliable. (05:33)

Some examples where weaknesses were found in products which allowed them to be fixed.

A major refrigerator manufacturer in the USA suffered from massive and expensive product warranty claims when a key component like a motor was burning out in fridges sold in hot states. This was found to be because the component was not able to cope with the higher temperatures and/or humidity in comparison to milder states.

Andrew also dealt with a company where the products were getting so hot and the internal fans were failing leading to product breakdowns. Although the product engineers had selected a fan that was able to withstand far hotter temperatures than the usual usage environment of room temperature, they failed to take into account that the transistors ran very hot and pushed the product’s temperature to over what the fans could operate in leading to them shutting down and the device overheating and failing. By finding what the product’s internal maximum internal temperature reached without a fan, it was possible to select a new fan that could operate at an even higher level (with some reliability margin) and then the issue was fixed. (14:30)

What kinds of methodologies can we follow to do test to failure?

Here are a few methodologies that might be used:

Doing a room temperature test and pushing the limits of the design just to see what happens. This can be done during the product design phase and you may choose to push the current and voltage higher than it’s meant to be which could give you an indication of possible failures.
Do Accelerated Life Testing (ALT) with humid, higher, and lower temperature conditions and you’ll learn what happens to the product under these environmental conditions. If no issues are found, more possible variables can be added to the same test such as lower or higher voltages that are increased or decreased incrementally.
High-stress testing helps you focus on the ability of the components to cope with excessive environments which could be too much voltage, current, high temperatures, high humidity, etc. If individual components fail it affects product reliability. In the aerospace industry, this kind of testing is critical because airplane components need to be able to handle repeated temperature and pressure changes for example, going from ambient ground temperatures to -50C at altitude, without failing.
Environmental testing provides peace of mind that the product has enough durability so it can handle extremes of temperature, humidity, vibration (during transport, for example), and more. This testing would be useful in the case of the defective refrigerators in hotter states example mentioned earlier.
Cyclical loading is when the product is tested under load and no load repeatedly and teaches you a lot about how components deal with stress. If these conditions aren’t tested by you, it will happen in the field. If components don’t stand up to the punishment, the result will be damaging and costly product returns and warranty claims.
Continuous monitoring of improvements shouldn’t be forgotten about, because if testing finds a problem and it’s fixed by engineers, but then they are too busy to monitor if the fix is truly stopping the same failures from occurring again over time by checking product returns for the issues in question, you won’t actually know if the fix has ‘stuck’ and is effective. Adding the testing and fix information to a ‘lessons learnt database’ could be useful here when monitoring improvements, as the database can be referred to easily and information is collected in one place. (19:56)

Products and industries where this is a very useful kind of testing to do.

Test to failure is commonly used in aerospace, automotive, consumer electronics, medical devices, and industrial equipment. If these tests are not done on products from these industries reliability risks would be very serious, for example, aircraft or cars crashing if components failed, or consumer electronics catching fire due to faulty batteries.

It must be mentioned that the company doing the testing needs to invest in samples that will be destroyed, but this cost is very likely to be lower than the cost of poor reliability which includes product warranty claims, returns, scrap, reputational damage, lawsuits, and more. (32:31)

Is test to failure wasteful?

As we’ve established, test to failure will result in scrap parts and products that are damaged and cannot be resold. But is this actually ‘waste?’ Waste has too many negative connotations for this testing and ‘useful’ is a better word, because it results in more reliable and durable products which will not need to be fixed, replaced, or cause more scrap due to repairs. This has a net result of fewer emissions, fewer resources needing to be used to make more components or products, fewer shipments of product returns, and less scrap going to landfills.

Another benefit is that once you have done this a few times you probably don’t need to do test to failure on future product iterations because you’re using components and designs that you know are going to be durable enough to result in a reliable product in customers’ hands, again leading to less waste. (37:20)

Don't miss a post