best youtube video on the subject ,well done and thanks a lot

Excellent! Thank you.

Simply neat & clear explanation. Thanks for taking the time to record this and sharing.

@TaylorSparks

4 жыл бұрын

Glad to do it and I'm glad it helped.

2,877 views and 25 likes???? Come on guys !!! more respect for valuable knowledge and the work for prepare this excellent video.

@TaylorSparks

6 жыл бұрын

Artur C. Freitas haha. Glad you find it helpful! Thanks for the promotion :)

Thank you my friend, it was really intuitive. I hope an example of the last part and how is that equation derived to understand the sigma of component.

this is exactly what I want to learn. I did't learn Weibull distribution in my probability course. I firstly face weigh bull in my wind resource processing. This distribution describe the uncertainty of wind speed.

@TaylorSparks

4 жыл бұрын

so glad to help!

Wow. I actually learned something. I went through so many documents online looking for a basic explanation. Couldn't find a single document with a clear explanation and easy maths. Thanks so much for sharing this! Now that I get it im curious to know what the limitations are of this technique, specifically the calculation of F. In what cases is does this analysis fail. And how do i know if a weibull analysis isnt representative of other samples or batches? Thanks again

@TaylorSparks

4 жыл бұрын

there are different ways to calculate the failure rate. The probabilistic failure wiki page is pretty good. Ultimately you should also look into how the weibull modulus confidence varies with number of samples. More samples gives you higher confidence in the weibu statistics. You can also look at the fit. The weibull statistics assume that your data can be fit the weibull equation, but if that's not the case in the analysis will not be useful.

@puffinjuice

4 жыл бұрын

@@TaylorSparks Thanks for the additional pointers. I'm going to apply this to the electrical breakdown strength of thin dielectric films. Thanks again. Have a great day!

Thank you for your amazing video! I am not an expert in math and I am struggling to understand how did you go from ln(σf)=0.8237 equal to 2.279 MPa (25:23). could please help me with that equation? Thank you!

@TaylorSparks

5 жыл бұрын

sure. to solve for sigma_f we take exp of both sides. exp(0.8237)=2.279 and that gets rid of the ln on the left hand side of the equation.

At 15:33, why did you multiply by a random number from 1-3?

@TaylorSparks

5 жыл бұрын

so, the only reason that I did that was because I was trying to generate two different data sets. The goal was to have one data set with a higher average value, but a wider variance, and the other data set with a lower average value but a smaller variance. The average number was determined by the random section, and the variance was determined by multiplying by a small random number. I only did this for instruction purposes because I didn't have a data set to work with so I created my own. when you are working on weibull modulus you will probably be given a dataset so you won't need to do anything like this to you or data.

Why do we need to subtract 0.5 from rank when calculating F?

@TaylorSparks

4 жыл бұрын

The 0.5 is necessary to make your small data set look more like reality. It's called median rank. You can read more herereliawiki.org/index.php/The_Weibull_Distribution essentially without the 0.5 consider this example. If you test 20 samples then when the 20th sample fails you would have F=20/20=100% But do you really think if you tested 1 million samples that none of them would outsurvive that 20th sample? Surely some would! Therefore, a 100% failure rate using your 20 samples is silly. Instead we use the 0.5 (or other approach for median rank). This gives us failure rate of F=(20-0.5)/20=97.5% failure. Therefore the failure rate is high, but not 100%.

What if you have several samples that fail on the same stress? If you rank them in order, you will get different probabilities of failure for basically the same stress!

@TaylorSparks

4 жыл бұрын

That's right

23:35 you can use the FORECAST formula in Excel to get that value at -13.8155 😉

@TaylorSparks

5 жыл бұрын

Great suggestion!

Very nicely explained. Weibull distribution is also used for composite materials like combination of fiber and matrix. But I want the weibull distribution for composite containing two different fibers and matrix. Please help me. Please give your email so that I can communicate with you.

@TaylorSparks

5 жыл бұрын

Sorry, I'm unfamiliar with Weibull distribution in composites.

@hiteshjariwala6150

5 жыл бұрын

@@TaylorSparks So many researchers have worked to find strength of composite material using two parameters weibull distribution. But I could not find it for hybride composite. Please try your best to help me.

Some guy is sleeping in the back