Reliability HotWire: eMagazine for the Reliability Professional
Reliability HotWire

Issue 52, June 2005

Hot Topics
Standard Based Reliability Prediction: Applicability and Usage to Augment RBDs

Part III: Completing the Picture: Using Lambda Predict to Augment your BlockSim RBDs

This is the third part of a three-part series about Standards based reliability prediction and ReliaSoft's newly introduced Lambda Predict software package. Part I introduced standards based reliability prediction and discussed when it is appropriate to use and when it should be avoided. Part II discussed the different major reliability prediction standards. This final part gives an example of how standards based reliability prediction and Lambda Predict can be used to complete a Reliability Block Diagram (RBD) when data is not available.

In this article, we study an example of an aircraft electronics parts manufacturer that is analyzing the reliability of a new design for a planes Radio Control unit, which consists of a Servos subsystem and a Receiver subsystem in a series configuration (reliability-wise). The design team is required to meet a reliability goal of R(50,000 hrs) = 0.99 before the design is given approval and a prototype based on this design can be produced.

Figure 1: The Radio Control unit

The manufacturer has tested the Servos subsystem in a previous application and determined its reliability model (Weibull with β=1.5, η=900,000 hrs) using Weibull++. However, the Receiver is a new concept design that the company has not manufactured yet. As a result, the company has no in-house tests or field failure data that describe the Receivers reliability distribution. Therefore, Lambda Predict will be used to perform standards based reliability prediction for the Receiver subsystem. The Receivers predicted reliability can then be incorporated into the BlockSim RBD described in Figure 1 to complete the reliability analysis of the Radio Control unit.

Data Entry in Lambda Predict
The following steps can be followed to create and analyze the Receiver subsystem in Lambda Predict.

Create a new project by clicking Create a New Project in the "What do you want to do?" window that may appear at startup, select New Project from the File menu or click the New Project icon in the MDI toolbar. In the Create New Project window, click Select standard and then select MIL-HDBK-217F, as shown next.

Figure 2: Create New Project Window

(Note: Although this example is presented using the MIL-217 standard, the procedures for using all other standards are similar; if the MIL-217 standard is not available on your computer, you can use any available standard that you find suitable, or contact ReliaSoft to obtain the missing standards/modules. The steps to follow in the example are the same but the results may be different if you choose a different standard.)

To add a block to the system, right-click the system in the System panel and select Add then Add Block from the shortcut menu. To add components to the block, right-click the block and select Add then Add Component from the shortcut menu. The Select Category window will appear, as shown next.

Figure 3: Select Category Window

For the purpose of this article, we will assume that the Receiver is made of 10 capacitors, 30 relays and 70 resistors. Select Capacitor from the list and either double-click it or click OK to add it to the block. The same procedure can be repeated to add the relays and resistors.

The components' properties are as follows:

The final configuration will look like the one shown next.

Figure 4: System Configuration
[Click to Enlarge]

The failure rate and MTBF results of the entire Receiver are displayed next to the MIL-217 project. The Receivers MTBF was determined to be 1,469,161.125 hrs and the failure rate was 0.68 fpmh. This predicted value can now be entered into the RBD defined in Figure 1. The exponential distribution is the only distribution that can be used to describe the Receiver, meaning we are assuming a constant failure rate (no early failures or wearout). This is one of disadvantages of using the standards based reliability prediction approach, in addition to not having actual data in order to quantify the reliability of this part (for more details, refer to Part I, which appeared in Hotwire Issue 50).

Now we are ready to analyze the whole Radio Control unit. Using BlockSims QCP, we find that R(50,000h)= 0.954, which is less than the specified goal of 0.99. Therefore, the design team should consider improving this design. Based on this analysis, the manufacturer has been able to make a prediction about a system that has not been built yet. The analysis gives a ballpark estimate of the inherent reliability of the design. In this case, a red flag was raised, pointing out that additional resources might be required for a reliability growth program. Regardless of the outcome of this analysis, a reliability validation test should always be performed; otherwise, the results can not be trusted. In other words, after building the prototype, it is highly recommended that the entire Radio Control unit or the Receiver be tested and reanalyzed with the help of Weibull ++ or ALTA in order to confirm the standards based reliability prediction results and improve the reliability estimation.

Copyright 2005 ReliaSoft Corporation, ALL RIGHTS RESERVED