I have a flange, originally 24" ANSI 300#, wich thickness PVElite declares insufficient. The design pressure is 2.16 MPa. The curious thing is that either the 150# and the 600# passed the calculations.

I prepared a file with screen shots of that, but I can´t realize how to add it to the post, even using Internet Explorer 7.

Can anyone tell me what is going on?

Filgueiras,

It looks like there may be an issue with the calculation. I have tried the same comparison and the 24" 300# is insufficient but the 20" 300# is OK. If you look at the calculation, the value of 'f' (hub stress correction factor) seems high for the 24" 300# flange.

Filgueiras,

Usually the issue is with the small end hub thickness. Try changing it to a thicker value and see what happens.

Stan, I checked the calculations, and found that PVElite choose f=1 for the 150# and 600#. For the flange 300#, it has made f=7.3

Scott, You're right, g0=12.7mm worked.

Thank you.

Now this is where it gets strange. If you reduce g0 it also works???

Good play, Stan!

g0=8.55mm also goes well.

I only calculated the flange at client´s request. So, I'm going to shake the dust off an old spreadsheet, and present it.

Regards,

I took some time this morning to type the 24inch 300lb flange problem into an Excel spreadsheet that was developed independently by an outside source. The stress results, moments, hub factors etc. were completely identical between PV Elite and the spreadsheet. This leads me to believe that there is some sort of spike in the equations for this combination of hub properties. As Stan mentioned, a smaller thickness also works.

Also note, the hub slope for this flange violated the 1 by 3 that ASME generally requires (see ASME Figure 2-4 sketch 6).

Scott,

Thank you, for expending your time in this question.

I'll check with the client about the reason why he requested the calculation of this flange, since UG-11/UG-44 say it's not required. I will not use a different thickness for g0, since the part is already ordered, and my practice is to have 100% match between drawings and calculations. But it's possible to consider the flange loose, in last case.

Anyway, this problem has been interesting and, for sure, improved my confidence in PVElite ( and in my spreadsheet, also .

Best regards,