Dear All,
I have noticed significant difference in result between manual calculation of flange leakage check as per ASME Sec VIII Div 1 and result from Caesar II for the same. Difference in flange stresses are very high when Peq is very high compared to design pressure and smaller size/rating flanges are more susceptible of this than higher size/rating flange. Though Peq is same in manual & Caesar calculation, but subsequently Caesar may be using Peq to calculate H and Mo in different way than manual calc (as per ASME code equation). I have highlighted this to Intergraph, waiting for their response. Meantime, anyone have any clue/experience about this, please share.
Regards,

I'm not sure, maybe the difference is how the gasket force (Hg) is considered.

You may note that ASME Sec VIII does not offer any clue about calculating Peq and/or handling Peq in calculation, so -in my opinion- you cannot say the "manual calculation" with Peq is strictly as per ASME code equation.

Best regards.

@mariog,
ASME Sec VIII div 1 doesn't say about Peq, but ASME B31.1 Appendix II gives deatailed guidance about it. I calculated Peq as per Kellog eqn and B31.1 guidance and compared with Caesar result, it is exactly same.

Sorry, I'm not sure I'm able to understand.

Peq is the equivalent pressure of external loads.
The problem is what you are doing with this Peq.

You may add it to the internal pressure, and the sum may be compared with ASME B16.5 flange rating or with other limit.
Or you can use the sum as "design pressure" in ASME Section VIII, Division 1
Or you would consider method modifications required by B31.1 II-4.2.3 Analysis of Flange.
Or maybe you would like to consider the approach of NC-3658.1(b) "the Design Pressure used for the calculation of H in Appendix XI shall be replaced by a flange design pressure =P+Peq".

I think that he said that in Caesar flange module there is a difference in results if you insert in their fields the axial force and bending moment or if you insert the Pd+Peq in the design pressure field as described in B31.1

Regards,

I think that he said that in Caesar flange module there is a difference in results if you insert in their fields the axial force and bending moment or if you insert the Pd+Peq in the design pressure field as described in B31.1

Regards,

I understand now.
I guess is something as explained here for PV Elite.
Probably it is the same approach in Caesar module.

I would add that for me it makes sense to follow NC-3658.1(b) "the Design Pressure used for the calculation of H […] shall be replaced by a flange design pressure =P+Peq", rather than consider P+Peq as design pressure.

The reason of my opinion is based on Kellogg book that mention “[…] analysis and experience which indicates that, with a properly preightened flange, the bolt load changes very little when a moment is applied, whereas the gasket loading changes appreciably.”
That means Kellogg team expectations were to have external loads transmitted to gasket loads rather than to affect W preightening load, and their Peq concept is a measure of this effect.

Indeed, including Peq only in H (or better said only in HD and consequently in H) means indirectly to consider
HG=W-H= W-PI/4*G^2*(P+ 16*M/PI/G^3+4*F/Pi/G^2)=W-PI/4*G^2*P-F-4*M/G
which is in line with Kellogg approach/ theory of plates.

Dear All,

I initially did not consider NC-3658.1(b), but subsequently consider the same though not agree with it fully. Still there is difference between manual and Ceasar result.
For example, for 6” 150# RF 1.1 flange, F=5kN, M=5kN-m
Result from manual calc, SH = 91.54 MPa where as from Caesar, 98MPa.

I have sent my calc to Intergraph & waiting their reply.

Regards,

I think that if you review the "flange.msg" that is located in the calculation folder you can easier spot the differences.

Regards,

Dear All,

Intergraph confirmed that my manual calc is correct. They are looking into Caesar flange leakage module to find out the reason for the discrepancy. FYI, I checked with ver 5.10, 5.20 & 5.30. But result is still the same. Looks like some bug there.

Regards,

Manoj,

Every day people use softwares to solve problems, but forget that is possible these same softwares may contain bugs. People like you, who cares about the true engineering, is always helping us with your knowlegde.

Congratulations.

Sincerely,

Miyamoto

Manoj,

We didn't say your hand calculations were correct. In fact, we found a couple of mistakes and informed you of it. The first mistake was in the computing of Hp where you used the 45 bar pressure instead of the 10 bar pressure. Then for your hydrostatic end force you again used the wrong pressure.

Then you re-did your computations and again we are looking into it. We aren't saying it isn't possible that there is a bug here, however I rather doubt it. When we get to the bottom of it, we will let you know the conclusion.

Loren,

I admit there was a mistake in my initial hand-calculation regarding pressure used for Hp & Hd and one of your colleague highlighted me the same. I revised my calculation accordingly but found difference in result again though lesser in magnitude.

My revised calculation again reviewed by your same colleague and emailed me saying 'I accept your calculations as being correct....', hence I wrote my previous message saying Intergraph confirmed that my manual calc is correct.

Anyway, it is good to know that you are rechecking my calculation. Appreciate if you highlight my mistake, if any, in the revised hand-calculation. I shall be happy if the conclusion is 'no bug in Caesar'.

Regards,

Here it is from Ray Delaforce, who handled this SR:

"In CAEASAR II there is a bolt correction factor. Now, on page 4 of my calculations I have computed the bolt correction factor which turned out to be 1.065. This bolt factor has to be applied to the stress equations for SR, SH and ST. Once the bolt correction factor is applied, your computed stresses will be the same as those computed by CAESAR II."

So, the conclusion is that there is no bug in CAESAR II. I am wondering if we should put all the equations in a single document and put it on our knowledge base for future users to check against their hand calculations. That is a huge amount of work you did... which is why we use computers for this stuff in the first place.

Loren,

It is indeed huge amount of work to solve so many equations manually, but we are in process of developing an excel spreadsheet for flange engineering (material selection, pressure rating, leakage, etc) for internal use in my present company. We bench marked CAESAR II for leakage calc, hence we performed manual calculation for a flange to compare Manual vs Excel result vs CAESAR II result. I sincerely thank Intergraph support for helping me in this process.

For benefit of this forum I would like to share the SR conclusion by Ray Delaforce about bolt factor as follows:

"1. The bolt spacing factor per ASME Section VIII, Division 1 is only mandatory for flanges in lethal service
2. The bolt spacing factor is more severe in ASME and than in TEMA (per CAESAR II)"

So, we can avoid using bolt factor for non-lethal service flange as per ASME, though CAESAR II uses bolt factor as per TEMA for all services. Since it worked fine for years, CAESAR II team may not like to include ASME provision. Loren, please correct me if you think otherwise.

Appreciate any suggestion/feed back from members of this forum.

Regards,

As I understand from the above discussion:

CAESAR II produces unnecessarily higher stress for non-lethal service flanges than ASME requirement. Let’s take Manoj’s finding as example, Sh without bolt factor = 91.54 MPa whereas from CAESAR II result (i.e with bolt factor) it is 98.0 MPa, almost 7% in this case. We all know how difficult it is to qualify flange leak requirement in some calculations and this un-necessary bolt factor in CAESAR II is making things miserable.

Moreover, as TEMA bolt factor is smaller than ASME bolt factor, CAESAR II result is less conservative for lethal service flanges. This is another discrepancy in CAESAR II result.

I wonder why not CAESAR II updates their programme to scrap TEMA and use ASME as default with an option to user to use or not to use bolt factor as per service.
Request a clarification from CAESAR II TEAM.

THANKS

Hi,

In reference to Loren Brown post ( #45525 - 10/27/11 02:56 PM ), was the manual calculation ever uploaded anywhere?

If they were, can someone please provide the download link for it?

Thanks.