Issue 51, May 2005

How are the values in the AVGOF and AVPLOT columns calculated in Weibull++'s Distribution Wizard?

The Distribution Wizard in Weibull++ 6 performs three tests. The values calculated for AVGOF are from the Kolmogorov-Smirnov (K-S) test, which is a goodness of fit calculation. More information on the K-S test can be found in many statistics books. As a rule of thumb, smaller values for K-S are generally better as they indicate that the difference between the points and the line are small. The equation solves for the difference (D) between the actual data points and the line represented by our parameters.

Results of the second test are reported as AVPLOT. This is a normalized correlation coefficient (rho). More information on the correlation coefficient is available at http://www.weibull.com/LifeDataWeb/least_squares.htm. The only difference is that in the Distribution Wizard, the correlation coefficient is normalized.

The third test is the Likelihood Value (LKV). More information on the Likelihood function can be found at http://www.weibull.com/hotwire/issue33/relbasics33.htm and http://www.weibull.com/LifeDataWeb/maximum_likelihood_estimation_appendix.htm.

Once these values are calculated, weights are assigned to each of the tests based on the parameter estimation method and a weighted average is presented and used to make the final decision on the distribution that best fits the data. For example, for MLE, a higher weight is given to the LKV test than to the rho test. For more information on the weights that are used, check the Advanced Wizard Setup.

How can I use Weibull++ to predict the optimum replacement time or the remaining life of a product?

You can use the Function Wizard in Weibull++ 6 to calculate the optimum replacement time. To do this, use Weibull++ 6 to obtain a failure distribution for your product. Next, insert a General Spreadsheet into the Data Folio by selecting Insert General Spreadsheet from the Folio menu. Select Function Wizard from the Tools menu. From the Function Wizard, select Table of block replacement times from the list of available functions. Enter the start time, end time and increment time, as well as the cost for the planned and unplanned replacement. (Please note that you cannot enter 0 for the start time. If the start time is 0, simply enter a small value, such as 0.00001.) Once you have entered all of the required inputs, click OK. A table will be created in the General Spreadsheet. To determine the optimum replacement time, look for the lowest value of cost per unit time in the table. Higher precision can be reached by narrowing the range for which the table is generated.

For more information on optimum replacement, see: http://www.ReliaSoft.com/newsletter/2Q2000/preventive.htm.

I have BlockSim version 6.2 and am trying to use container blocks and mirrored blocks; however, the options to add these blocks are not available to me. How can I add container blocks and mirrored blocks to my diagrams?

The options to add container blocks and mirrored blocks are not available to you because the application is set to the Simple setting. In the Simple setting, only the options for performing basic system analysis, which does not include container or mirrored blocks, are available. To switch to the Advanced setting, which allows you to perform advanced system analysis using container and mirrored blocks, select User Setup from the File menu. In the User Preferences window, click the Other tab and click the Modify Default Presets button. In the Default Settings for MyBlockSim window that appears, click the Advanced option and click OK. You will now be able to add container blocks and mirrored blocks to the Diagram Sheet.
 

Copyright © 2005 ReliaSoft Corporation, ALL RIGHTS RESERVED

Comments or Questions: Webmaster@ReliaSoft.com
weibull.com is a service of ReliaSoft Corporation
Copyright © 1992-2008 ReliaSoft Corporation, All Rights Reserved
Last updated: 06/26/2008 06:58:59 PM

SITE [Home] [Site Contents]
LEGAL [Copyrights & Notices]
[Terms of Use] [Linking Guidelines]
[On-Line Privacy Statement]