EN 13445-3 flat bottom weld detail and fatigue

[MGZmechanical]

Hi, we're designing a pressure vessel according EN 13445. We've to do some flat bottom weldig and we're looking at EN 13445-3 Annex A Table A-3 details E19 and E25. I'm a bit surprised that the throat thicknesses are:
- for E19 a = 0,7*e on each side
- for E25 a = e on each side.

I think that's a bit of over-welding. Is it possible to use less weld?

On the other hand. I have a copy of AD-Merkblatt 2000 (S2 Annex 3) which comes with an example of a calculation of a weld under fatigue. Material P265GH (265 MPa yield strentgh, 410 MPa ultimate strength). The weld detail is class K1 (class 63 in EN 13445) at 250 000 cycles.

It says:
2*sigma = pseudo elastic stress range
2*sigma = (5*10^11/250 000)^(1/3) = 126 MPa

That's fine. Now if I take at 5000 cycles:
2*sigma = (5*10^11 / 5000)^(1/3)=464 MPa That's bigger than the ultimate strenght.

When I studied fatigue loadings (Shigley's Mechanical engineering design) I used to consider a straight line from 0,9*ultimate strength at 10^3 cycles to 0,5*ultimate strenght at 10^6 cycles.
With the EN method the slope is constant (1/3) so it's independant of the material (I understand that due to residual stresses and cracks in welding) but they put no limit to upper strength and they say you must consider fatigue over 500 cycles.

Any ideas?

 

[corus]

I think the line you have drawn 0,9*ultimate strength at 10^3 cycles to 0,5*ultimate strenght at 10^6 cycles must represent the line to which the material with stresses below those values will have 'infinite' life. It's not a limit on upper strength, as such, as the component may still fail under static loads.
For material with welds, the material strength doesn't come into it as you are no longer looking at crack initiation, but crack growth, ie. it's assumed that the weld includes defects/cracks already. The limit on the 'stress range' will be governed by the weld classification. Remeber also that the stress range can of course be greater than the UTS of the material.
For a limit on the strength of the component you'd have to look at primary/seoncdary stress limits for the mechanical/thermal load history.

Tata

 

[MGZmechanical]

Hi, one thing I forgot to upload the weld details.

Regards,