DFMEA Occurrence Rating

Quick reference

• 00:04 Hi, I'm Ray Sheen, we'll continue at step five,
• 00:08 scoring in the Design FMEA by scoring the probability of occurrence.
• 00:14 The occurrence rating is the probability of occurrence of the failure mode.
• 00:18 And what we are determining is the probability of whether that
• 00:21 failure mode is likely to occur in the expected life of the product.
• 00:25 If the likelihood that this type of failure would occur to
• 00:28 most customers is high, then the score is high.
• 00:32 And if the likelihood that this type of failure
• 00:34 will occur to most customers is low, the score is low.
• 00:38 Now in order to determine the likelihood, you first have to determine the reason or
• 00:42 cause for the failure to occur.
• 00:44 And that's what we will actually be listing in the Design FMEA form, and
• 00:48 that is what we will be scoring.
• 00:50 Keep in mind that it is very common for a particular failure mode to occur for
• 00:54 multiple causes.
• 00:56 When that is the case, each cause becomes a separate line in the Design FMEA form.
• 01:01 And of course, each cause will be scored independently of the other causes.
• 01:06 Now to determine that probability of occurrence score you should be using
• 01:09 historic data.
• 01:11 So let's take a minute to review the appropriate sources of data.
• 01:14 As we consider data sources,
• 01:16 you want them to be as relevant as possible, so if there was any
• 01:19 actual history on this product based upon your development testing, use it.
• 01:24 Keep in mind, if this is a new product, there is no data on this product.
• 01:28 But there may be some on other similar products.
• 01:32 There could also be external or industry data for other similar products.
• 01:36 The science and discipline of reliability engineering
• 01:39 has made some tremendous progress with virtual twins and simulation.
• 01:43 So that might be a useful way of predicting failures.
• 01:46 And for source components, the supplier or
• 01:49 manufacturer of the component may be able to provide failure data that you can use.
• 01:54 Regardless of the source of the data, be sure you
• 01:56 are scoring the probability of the causes of the failure, not the failure itself.
• 02:02 As I mentioned in another lesson, there are many standards and scales in use.
• 02:06 I've selected three of the most popular ones for doing this assessment.
• 02:10 One is based upon the measure of the percentage of the customers
• 02:13 who have had this failure.
• 02:15 Another is based upon the design history of the product that we are developing.
• 02:19 And the third is based upon the knowledge and
• 02:21 experience with the application in which the product will be used.
• 02:25 Use the percentage if that's known, otherwise,
• 02:28 use the score that is the highest of the other two.
• 02:31 A score of 10 is when the failure mode is almost certain or
• 02:34 greater than 90% of the time a customer experiences this cause of the failure.
• 02:39 From a design history standpoint, we have no history, so
• 02:42 we have to treat it as high.
• 02:44 No other product or technology like this has been used.
• 02:47 And from a product application standpoint,
• 02:49 we know very little about the application and none of that from direct experience.
• 02:55 If the percentage is 80 to 90% or if we know something about the design history,
• 02:59 we can go to a score of 9.
• 03:01 But of course, we use this score of what we know is that everyone has this problem.
• 03:06 Similar products have experienced this,
• 03:07 and we must put some controls in place to account for it.
• 03:11 From an application perspective,
• 03:12 there is very little experience with the application.
• 03:15 It isn't totally new but
• 03:17 it's not something that we have been doing on a consistent basis.
• 03:20 The score of 8 is still high with the probability of 70 to 80%.
• 03:25 We have limited experience with this type of design, and
• 03:28 if we had experience, it tells us that failures are likely to occur
• 03:32 based upon the operational conditions for the product.
• 03:35 We have a little familiarity with the application, but only a little.
• 03:40 Now let's moved to a score of 7, this is a moderately high percentage of 60% to 70%.
• 03:46 From a design history, we're copying design of a developed product but
• 03:51 we're using it in a very different way.
• 03:53 So the product works in the current application and
• 03:56 hopefully will work in another one as well.
• 03:58 From the application standpoint,
• 04:00 our problem is that we don't know any of the current customers, so
• 04:03 there is no way to get direct, current knowledge of the application.
• 04:07 A score of 6 is for a failure more with a probability of occurrence of 50 to 60%.
• 04:13 We know from a design history standpoint that the design will work, but
• 04:17 some of the similar designs have needed process controls to protect it
• 04:20 from this failure mode.
• 04:22 With respect to this application, we are somewhat familiar with it.
• 04:26 A score of 5 is used when the probability is now below 50% but greater than 40%.
• 04:31 We've taken a known design and copied it for
• 04:34 the start of this development but then we've made some significant changes.
• 04:39 This is often the category we use with major product upgrades.
• 04:43 And with respect to the application, we have made products for
• 04:46 this application in the past but we don't currently have any in production.
• 04:50 Now let's look at the low score of 4,
• 04:52 which is a probability of 30 to 40% of the customers experiencing
• 04:56 this cause some time during the expected life of the product.
• 05:00 From a design history standpoint,
• 05:02 we know there have been failures of this type on similar products.
• 05:05 From an application standpoint,
• 05:07 we know some of the customers that this part is targeting.
• 05:11 A score of 3 is very low,
• 05:12 with only 20 to 30% of the customers ever seeing this cause.
• 05:16 No similar products have ever needed to put in process controls to watch out for
• 05:20 this type of failure, and
• 05:21 from an application perspective, we know this application well.
• 05:25 By the time we get to a score of 2, things are starting to become pretty remote.
• 05:29 Only 10 to 20% of the customers have ever seen this type of problem.
• 05:33 With respect to the design history we have copied this new product from an existing
• 05:38 product with no significant changes, think of it as a product line extension.
• 05:43 And from an application standpoint, we not only know the application,
• 05:46 we have many products that are already serving this application.
• 05:51 Finally the score of 1,
• 05:52 less than 10% of customers have ever seen this failure cause.
• 05:56 We may have eliminated the failure mode to preventive design techniques like
• 06:00 Poka Yoke.
• 06:01 And from an application standpoint, we know this application inside out.
• 06:06 So let's score the ball of our ballpoint pen example and see this in action.
• 06:11 The reason the ball may fall out is incorrect assembly and
• 06:14 since this is not our first pen, we have a great deal of experience with this ball
• 06:18 assembly and can score it as a 2.
• 06:21 The next failure is if the ball stops rolling, the cause of this is that the ink
• 06:25 dries up and acts like a glue holding the ball to the ink tube.
• 06:30 This has bumped up to a 3 because a slightly different formula of ink is
• 06:33 being used.
• 06:35 Our experience tells us that there is not a problem but
• 06:38 it is not an exact copy, so we go up to a 3.
• 06:42 With respect to the failure of ball corrosion,
• 06:44 the cause is that the pen is being used in a corrosive environment,
• 06:48 it is exposed to factors it will create the corrosion.
• 06:52 This one jumped to five.
• 06:53 This product is being developed for businesses that support the offshore
• 06:56 drilling industry and the environment is definitely corrosive.
• 07:00 Some of the products we have sold to this industry in the past have
• 07:03 experienced this problem.
• 07:05 With respect to the wearing of a flat spot,
• 07:07 that is based upon the pen being dropped or damaged.
• 07:10 Which causes an initial nick in the ball
• 07:13 that then grows into a flat spot through wear and tear.
• 07:16 Now we have only seen a few isolated examples of this problem before with
• 07:21 products built for this industry, so we score it as a 4.
• 07:25 The last failure is the ball skips while rolling, and the reason we have assigned
• 07:29 for that is one of contamination causing problems with the ball alignment.
• 07:34 Well in this case, there are no similar designs that have needed any control for
• 07:39 this problem, so we score it as a 3.
• 07:41 You can see why it's so important to have subject matter experts who know
• 07:45 the product technology and customer application on your FMEA analysis team.
• 07:49 You need to draw from their expertise to score the probability of occurrence.
• 07:54 Okay, that's our second score, the probability of occurrence score.
• 07:59 That's one that is most likely to change during the life cycle of the product as
• 08:03 you gain experience with the application.

Lesson notes are only available for subscribers.