Dear All concern

as per the subject , in one of the project we wanted to use H SECTION inplace of trunnions, so what are the necessary changes in KELLOGS formula

also i am not able to understand the line load concept as explained in kellogs when it comes to using H SECTION

also when we use H SECTION at elbows what can be the relevant change in keloggs formula

please guide me , i already passed my confirmation for WRC but we want to avoid WRC and want to use kellogs formulae

Hi,

We have used this. We were concern about water getting into the trunnion, on a special application. The moment of inertia is totally different form a circular section, hence the stresses.
If you what to check only the trunnion use a H section with almost the same inertia of a circular pipe……in that way the trunnion should be ok ( quick way).
If you want to check the header (you should), you need to consider the linear load of a H section and then calculate stress on the main header….is exactly the same, what it changes is the geometry.
Hope it helps.

Thanks OG

but what can be the line loads of a H section , i mean to say what can be the terms which can replace denominators in KELLOGGS FORMULAE for Line Loads calculation ,

can you provide me any sample calculation or any formula that shall help me to quickly grasp the matter

IMHO the Kellogg formula cannot be applied to an H section trunnion attached to a pipe. The local stress introduced by the 'sharp ends'of the beam section may cause high local stresses that would not be predicted by Kellogg. It is a closer situation to the horn of a vessel saddle and thus a Zick type solution.

If a substantial pad were employed, the Kellogg approach might be applied using the perimeter of the pad to calculate the line load.

As a collateral discussion, is someone ready to share us the hypotheses behind the (in)famous Kellog formula? I don’t refer to how to apply it or why…
How it was simulated the connection trunnion-pipe?
Maybe it will be a good opportunity to stop calling this formula everywhere.

Regards,

There are some indirect evidences that the work of Kellog Co has been made by considering a linear distributed load inside the trunnion (probably a square surface?) on a cylindrical header.
I guess longitudinally there was a "bending of beam on elastic foundation" model,transversally there was an uniform circumferential radial line load applied on cylindrical shell, all calculated to balance the actual loads.

I understand Kellog formula was contemporary and has been influenced by Bijlaard’s work, being intended as a "simplified approach with a reasonable understanding of the extent of his deviation from more accurate solution". Bijlaard also considered such model but together with simplified cylindrical shell equations.

In fact I don’t know the Kellog’s Co old work and I’d like to appreciate this only as an historical, rough engineering acceptable tool to handle the problem.
It would be nice to know deeply Kellog work, after that… it should be all.

Anyway, to expand today this approach to a trunnion-on-elbow (where there is a torus shell) and/or to a special-shaped trunnion (H-section) it would get loose additional serious percents to an already inaccurate model (maybe adding more 15-20% inaccuracy to an already inaccurate model?... or perhaps the errors may compensate…I don’t know, a FEA analysis might reveal this, if one is interested…).

I understand the attraction for a formula; psychologically an engineer would be more comfortable having a formula instead of a FEA result. From this perspective also stress calculation by Caesar as FEA improvement was also a big step as mentality!
Only a question: why are you comfortable with Caesar results and are you still calling Kellog formula?

Mariog, There is a brief explanation of the method in Kellogg's Design of Piping Systems, page 85. FE studies have shown results to be be reasonable where the trunnion and pipe diameters are similar. IMHO it is not really calculating stresses, rather it is a numerical comparison to designs known to be acceptable due to operating history. Arguably, the trunnion on elbow arrangement has a far greater line length and should thus be more robust. Stress in the acute crotch may be a problem with thin wall ,locally flexible arrangements.

Moverz, you are right, of course. And of course I understand what you are saying and (I hope) more than.
But I cannot feel comfortable with such model approach extrapolated beyond its own limits. In our days we have better tools.

Regards