SIF Calculation for 45 deg Tee with Gussets 2

[datsnl]

Guyz,

I am doing Stress Analysis of a DN1050 dia pipe which has a 45deg Tee. The pipe already exists on the site and this is like to like replacement. This tee has 4 gussets placed 90 deg to eachother. The tee fails in Caesar II analysis because I dont know how to model gussets in Caesar II. I know that I can put my own calculated SIF at the tee, but I dont know how to calculate SIF for tee with Gussets. I know that Calcualtion for Reinforced tee is available in B31.1 but not for Gussets.

Thanks

Dharmit

 

[richay]

Piping Flexibility programs (CAESAR II included) use "3D Beam Elements" to model the piping system. You can think of these as "infinitely thin sticks". That's it, you can't model anything else. So therefore you can't model gussets.

The piping codes only address SIFs for 90 deg tees. Reinforcing is in the form of a pad only. If you deviate from these restrictions, you are outside the scope of the piping codes. That leaves you with the options of (a) acquiring the necessary data from a vendor, or (b) deriving the data yourself via FEA.

Richard Ay
COADE, Inc.

 

[datsnl]

Thanks richay,

I think now its the job of our FEA guyz to sort this out.

Cheers

 

[JohnBreen]

datsnl

The B31 Codes address reinforcing pads only. There is good reason for this. It had been proven that gussets are of no value in reinforcing branch connections for internal pressure. The famous Kellogg book has an illustration in which they show several types of branch reinforcement that were in use at the time the book was originally published. The text provides comparative numbers in terms of percents of relative strength compared to a B16.9 Tee (these data came from "burst tests). The gusset type reinforcing schemes are shown to be of very little improvement over the unreinforced fabricated branch connection.

We have done (for several NASA research centers) a significant amount of FEA of 90 degree fabricated branch connections and of lateral (angled) fabricated branch connections up to 45 degrees, for large diameter gas and air piping. The analyses of fabricated branch connections with triangular gussets spaced at 90 degree locations and at 45 degree locations show that the gussets will yield under small to moderate bending moments (moment at the intersection of the branch and run pipes). Also there is considerable punching shear introduced to the run pipe by such gussets when the moment loading occures. One the gussets yield, they are no longer providing any structural reinforcement and they are introducing unwanted "side issues" like shear.

I would suggest that you write a management of change (MOC) report to allow you to abandon the gusset arrangement in favor of the more predictable pad type reinforcement illustrated in the B31 Codes. With documentation showing a ratioinal assessment it can still be a "like-kind" replacement and you will have the benifit of conventional fabrication with all the years of satisfactory service that it brings with it. As Rich Ay points out, the fatigues tests of piping components that our ASME B31 SIF's are based upon did not include lateral fabricated branch connections so the SIF's were never established for this type of construction. Some companies increase the 90 degree SIF's by 50 percent for use with lateral fabricated branch connections. I would also suggest that you do the "area replacement" calculations to be sure that your reinforcing pads include adequate reinforcing metal (look at Appendix "H" of B31.3). Some old systems were not originally hydrostatically tested and it is not unusual that when the hydo-test is done after the installation of new branch connections, there are some "surprises".

Regards, John.

 

[NozzleTwister]

Thanks for your valuable post John.

While I realize that gussets are of no value in reinforcing branch connections for internal pressure, I did think that they would help prevent bending at the branch intersection. Gusset yeilding and punching shear never crossed my mind, thanks for the those comments.

For laterals, unless the branch is small compared to the header, instead of adding a re-pad, I usually specify a heavier wall thickness for the run and the branch to accomodate the area replacement. For the length dimensions of the thickened section I generally use catalog dimensions from Taylor Forge or other manufacturer of fabricated laterals and taper bore the ends to match my pipe. This gives the shop the option to fabricate the lateral themselves or purchase it.

NozzleTwister
Houston, Texas

 

[datsnl]

Thanks John,

I had never thought that gussets would prove to be a hindrance. I will discuss this issue with our Mechanical Head and also do the calculations for the reinforcing pad as per Appendix "H" B31.3 that you have suggested.

Thanks

"Thought we might have best computers and the softwares in modern times. We dont have the experience that is stored behind the grey hair and years of field work."