Talk:Rainflow-counting algorithm

Mistake in table

I may have done a mistake myself but the results in the table of the example seems wrong. I am caculating these results instead : (W for Whole, H for Half)

10 (2W) ; 13 (2H) ; 14 (1H) ; 16 (1W 1H) ; 17 (1H) ; 19 (2H) ; 20 (1H) ; 24 (1H)

It doesn't seem to have place for any 29 or 22 for example. Frsix (talk) 15:43, 27 August 2021 (UTC)[reply]

Reference to Sandia code

It is irrelevant to mention the Sandia National Laboratories LIFE2 code. It sounds as publicity.

Agreed. It has been removed -NeedsGlasses (talk) 14:17, 16 July 2019 (UTC)[reply]

Mistake in Figure 4?

I'm by no means an expert in this field, but from reading the algorithms it seems to me that the topmost blue half cycle in figure 4 (starting at around value -14) should "water drop" downwards until it reaches the righmost peak at value +15. This, in turn would influence the some later troughs. If this is no error, please help me correct my understanding of the algorithm. Bleistiftspitzer (talk) 14:16, 6 November 2012 (UTC)[reply]

I agree with you, the figure 4 needs to be corrected. --Irmo322 (talk) 19:58, 20 December 2018 (UTC)[reply]

Reply regarding the "mistake"

As the flow merges with the following flow (starting about 0) I think it flows opposite a tensile peak of greater magnitude. If you use the new flow it merges with as a reference this is true for Figure 4. But this is just my understanding of it and I'm no expert on the matter. — Preceding unsigned comment added by 46.235.154.16 (talk) 13:06, 13 December 2012 (UTC)[reply]

I'm not an expert too. But the example C of figure 3 is a counter example of your argument. --Irmo322 (talk) 19:58, 20 December 2018 (UTC)[reply]

"If you use the new flow it merges with as a reference" - you don't, though. The new flow is the one that terminates as it merges with the older one. The wikipedia page itself states this.
Moreover, imagine if the stress history started at -15 (location of the (A) in Figure 3) and terminated at the same value: Half-cycle 5 would continue down to -15. But there would be no matching cycle among the compressive valleys. This is incorrect according to https://www.sciencedirect.com/topics/engineering/rainflow-counting - "When a loading history is periodic, the loading history needs to be rearranged to start from the largest extremum point and this extremum point is repeated at the end, in effect closing the largest hysteresis loop. All inner reversals therefore pair up to form cycles." I would thus definitely agree that the image is wrong. --User5849 (talk) 10:00, 13 September 2019 (UTC)[reply]

Hello, i generated two new figures (3 and 4) that should correct the errors. While generating the new images, i also tried to use the same values as in the existing example. When evaluating this profile, i also noticed, that the results presented in the table seem to be not correct. I will update the figures, and i think i will present the suggested table values here for a few days before changing the values in the article. — Preceding unsigned comment added by Franzxf5 (talk • contribs) 18:14, 11 March 2020 (UTC)[reply]

Link to German page

https://de.wikipedia.org/wiki/Rainflow-Z%C3%A4hlung

is the German Version of this topic. I was not able to link it, could anyone please do that?

Geschichte09 (talk) 11:50, 18 November 2014 (UTC)[reply]

Most popular as of 2008?

Any data to support that claim? — Preceding unsigned comment added by 104.129.196.110 (talk) 16:18, 7 July 2016 (UTC)[reply]

It has been removed. It was unsupportable and not very informative. NeedsGlasses (talk) 01:04, 15 July 2019 (UTC)[reply]