Weibull Probability Plotting of complete data using median ranks with example

This video explains step-by-step procedure for probability plotting of failure data. Probability plotting is a technique used to determine whether given data of failures follows a distribution. The video also explains various types of failure data: complete, right censored, left censored and interval censored data. Viewers can try Weibull Probability Plotting to estimate shape parameter beta and scale parameter Eta. The video also shows how to estimate median rank for complete data.

[@gajananshiralkar9670]

This is best video for weibull in you tube.

[@instituteofqualityandrelia7902]

Wow! I am delighted to know! Thanks for appreciation.

[@candycandy319]

Thank you! finally get how all these come together

[@instituteofqualityandrelia7902]

Thank you! I am glad to know that you found it useful!

[@ranjanpal7217]

Very nice video...CAn we estimate weibull parameters by Python?

[@uhemant1]

Thanks. I am not a Python user. If I learn, I will try. But I am sure, it is possible.

[@sridharsdrawingbook6316]

Very very useful sir. Not missed out a single point. The teaching flow is amazing for beginners. Thank you so much.

[@instituteofqualityandrelia7902]

So nice of you! Thanks.

[@seshansesha7645]

Thank you

[@instituteofqualityandrelia7902]

My pleasure!

[@ArjunSharma-wi3jp]

HI SIR , IW ANT TO KNOW WHY DO PEOPLE PLOT LN(LM 1-/1-F(x))ON X AXIS AND GET THE MODULUS

[@instituteofqualityandrelia7902]

Weibull equation has power. So we take a double logarithm to make it linear.

[@sangram0403]

Very nice video. Explained well about the weibull plotting and finding shape parameter.

[@instituteofqualityandrelia7902]

Appreciate your feedback! Thank you!

[@jonmoore8995]

Very good. Please also do video of theory behind median rank.

[@instituteofqualityandrelia7902]

Thank you John. I will try it sometime in future. Appreciate your input.

[@tensioattivo86]

I cannot understand last logarithm plot... and method to calculate beta

[@instituteofqualityandrelia7902]

The x-axis of the probability plot is logarithm of time t and the leftmost point is 10. Also, to determine beta i. e. shape parameter, you need to draw a parallel line to the plotted line passing through the 63.2% point as shown and read the beta parameter value from the scale shown at the top.

[@mohamedabdelrahman169]

Very useful video... thank you

[@instituteofqualityandrelia7902]

Thank you!🙏

[@joshuagodwins5386]

thank you for the explanation! I'm curious what to do if we get a data set where some of the items are still running, do you exclude them from the data as they haven't failed or do you carry on as usual

[@instituteofqualityandrelia7902]

You should include it as censored data. If an item is running, it is right censored data.

[@alinecristinapin]

Estou aprendendo a manusear o programa. Estou com algumas dúvidas .Tenho um modelo de exemplo na qual não possuo os valores para ser inserido somente o valor de Beta, n e Y. No modelo pede para utilizar o modelo Weibull Mista 3 subpop. Possui uma fórmula também. Se optar pela opção Definida pelo usuário consigo inserir a equação, porém não localizo a opção de inserir os meus parâmetros Beta, n e Y.

[@yfc631]

All heroes are not wearing cap.. 👏👏👏👍

[@instituteofqualityandrelia7902]

😊

[@titox7754]

Thank you very much! It was extremely useful!

[@instituteofqualityandrelia7902]

I am you liked it!

[@md.hazratali6796]

good job! can you suggest any good explanatory reference book ?

[@instituteofqualityandrelia7902]

Thank you! There are many good reference books. The one that Ilike is by Patrick O'Connor. Ebelling's book is also good.

[@sridharsdrawingbook6316]

Sir,
Can we use the minitab version for accelerated testing and life estimation of explosive sample?
Example:-
Different samples of same explosive fail at different times at constant accelerated temperature. Can I use Arrhenius weibull distribution to predict their life at usable temperature?
One of your lectures well predicted the life of other components at various temperatures. Request guidance and suggestions in case of explosives.
Thank you so much in advance.
Regards

[@instituteofqualityandrelia7902]

Thanks for asking a very good and interesting question. The answer needs to be evolved. You need to define what is a failure. If this is predominantly result of temperature, Arrhenius model could be appropriate. If there is some other stress, Eyring Model may be better. Also, analysis is possible if you are doing testing at three different temperatures. Please write to me on [e-mail] for more detailed discussion.

[@sridharsdrawingbook6316]

@@instituteofqualityandrelia7902 Thank you so much sir. I've set the similar failure criteria at three different temperatures for ALT much higher than usage temperature . I followed your tutorial on MINITAB. I'll wait till all the samples fail.
I wanted to try this method as not much literature available on predicting life of explosive samples using this method.
Thank you for the postive thought and push. It will be very interesting if the results are compared with traditional methods.
I would forward the results in mail. Request your guidance whenever free for any corrections in procedure and assumptions. It would be a great help.
Regards,
Sridhar

[@momoabdul2755]

Thank you for making this video.

[@instituteofqualityandrelia7902]

Appreciate your feedback!

[@drpyne4628]

Thanks! Nice presentation

[@instituteofqualityandrelia7902]

Appreciate your feedback Dr.Pyne!

[@ramachandramurthyjakkula7561]

TQ VERY MUCH FOR THE VIDEO

[@instituteofqualityandrelia7902]

Welcome! Appreciate your feedback!

[@mohkarim4384]

thank you sir please
please how to estimate beta and alph i have parameter of weibull
axe x=TBF=1907
,1913,1933,1949,1964,1980,2028,2030,2039,2069,2092 with hour (h)
axe y= F(i) median rank% = 15,13,20,28,35,43,50,57,65,72,80,87,95

[@instituteofqualityandrelia7902]

Thanks for your keen interest! You sample size is 11. I assume that all are failures, so this is a complete data. So the median ranks that I see are 6.1, 14.8......., 93.9. The median ranks indicated by you seem to be incorrect. See link www.weibull.com/GPaper/ranks10_13.htm. If you fit two-parameter Weibull, the shape parameter is 36.8. However, three prameter Weibull is probably better with shape about 1.4, scale 117, and Failure free life as 1888. I checked this on Minitab with LSXY. Email me for more discussion on [e-mail].

[@sunnyj1967]

This simply brilliant.

[@uhemant1]

Thanks!

[@snehalpatil8788]

Thanks for video !

[@instituteofqualityandrelia7902]

Welcome Snehal!

[@TheSanthoshteja]

Thank you so much Sir

[@instituteofqualityandrelia7902]

Appreciate your feedback!

[@orhansargul5000]

ı need to ask you a question

[@instituteofqualityandrelia7902]

Watch my video th-cam.com/video/WR4-A_zVK28/w-d-xo.html

[@mohammedali7335]

HOW CAN WE DIFFERENTIATE WEIBULL DISTRIBUTION ,NORMAL,AND EXPONENTIAL FROM GIVEN RELIABILITY DATA?

[@instituteofqualityandrelia7902]

Watch our other videos for this:
th-cam.com/video/JuQ_AEtj1vE/w-d-xo.html
th-cam.com/video/wBG9POVWjRc/w-d-xo.html

[@MrNikhilshenoy]

Brilliant explanation!

[@instituteofqualityandrelia7902]

Glad you liked it!

[@gyanchandrakushwaha8604]

Great Explanation

[@instituteofqualityandrelia7902]

Thanks a lot! Appreciate your feedback!

[@sudhansugrahacharya7094]

Sir how can we apply it in estimating three parameter weibull distribution

[@instituteofqualityandrelia7902]

You will see some curvature in the graph. Then you need to try some values of Delta. Else use software.

[@onlytruth9321]

Sir what are the advantages of different kind of distributions? (But we already know that for variable failure rate we need Weibull and for constant failure rate e need exponential distribution. ) is there any other advantages??

[@instituteofqualityandrelia7902]

Hello 'Only Truth',
We need to choose distribution that fits our data. It is like a shirt. If I need XL size, I should not buy L size. Both shirts are good for persons whom these fit well! Also, remember that Exponential is a special case of Weibull with beta=1.

[@yfc631]

Sir u just osm

[@instituteofqualityandrelia7902]

Thank you Ismail! Appreciate!

[@mohabelshazly5653]

Thank you very much for the Video, very helpful, i have a Question regarding this, i have set of data of replacement times only for main components of a System, can i apply MRR to These data in order to calculate the weibull Parameters? thank you in Advance.

[@instituteofqualityandrelia7902]

I suppose yes, you can. What is MRR? Are you receding to median ranks?

[@shiguoliu7914]

if two failure have the same failure cycles, that meant there are two failure have same rank, how to handle this case? thanks

[@instituteofqualityandrelia7902]

You will have to rank these one after another.

[@vaibhavidige6869]

Thank you sir for this video. Can you please help me out for plotting weibull curves in excel based on service life prediction of building data. Plotting of Condition index vs.Time in years

[@instituteofqualityandrelia7902]

Hello Vaibhavi Dige, Appreciate your interest!
Please inform more details of your problem and data to me. My email id is [e-mail].
Good luck!
Hemant

[@ArunKumar-eu4sc]

Is it required to rank the censored data points also.. my sample have many components still running in the field ?

[@instituteofqualityandrelia7902]

Thanks for your question and interest. The censored data is used to sort the data. However, its median rank is not calculated and plotted. But it changes the median rank. In this video, I have explained Weibull plotting for only complete data. Apologise for some delay in my reply.

[@candycandy319]

very helpful

[@instituteofqualityandrelia7902]

Thanks! I am glad you found it helpful!

[@nicolaswiles3236]

Very helpful, thanks!

[@instituteofqualityandrelia7902]

Thanks for your feedback! I am glad you found it helpful!

[@instituteofqualityandrelia7902]

Thanks a lot! Appreciate your feedback!

[@alaeddinesel7891]

I need a program that defines parameter weibull at paper weibull

[@instituteofqualityandrelia7902]

I have not understood your question. Pleasee clarify. Thanks.

[@dartagnandebatz3304]

So well explained Mr. Urdhwareshe!, thank you so much.

[@instituteofqualityandrelia7902]

Most welcome!

[@mukulpandey8445]

Hi very nice explanation. I am interested to on failure forecast using Weibull, the data set i have got have different failure mode (electrical/ mechanical) of a pump failure. how can we combine them together to get a Weibull distribution for system failure and then use it for failure forecast? appreciate your insight on the problem

[@instituteofqualityandrelia7902]

Different failure modes should not be combined and analysed separately for each failure mode. Minitab supports analysing multiple failure modes in single command. In case you need help, please send email on [e-mail].

[@fourierrocket]

Thank you for thia video.
It answered all of my questions on how to perform Weibull plotting.
Best Regards

[@instituteofqualityandrelia7902]

Glad it was helpful!

[@ruchirshrivastava6638]

Excellent video sir. True Help.

[@instituteofqualityandrelia7902]

I am glad it helped! My best wishes!

[@dartagnandebatz3304]

So, so well explained, thank so much !!!

[@instituteofqualityandrelia7902]

Glad it was helpful!

[@chekradoua]

thank you for all you presentation

[@instituteofqualityandrelia7902]

My pleasure!

[@mutiardini]

thank you for your explanation

[@instituteofqualityandrelia7902]

Glad it was helpful!

[@luciandersonmarquesferreir1165]

Thanks Mr. Hemant.

[@instituteofqualityandrelia7902]

Most welcome!

[@parviz1961]

Thank you very much!

[@instituteofqualityandrelia7902]

You're welcome!

[@alaaisraa9459]

Thank you for this video, please I have a question, how do we estimate the failure function, if I have the number of data exactly 20

[@instituteofqualityandrelia7902]

I believe you mean the median rank for 20. You can either use median rank tables for n=20 or use the formula (j-0.3)/(n+0.4) as shown in the video.