How to model a trunnion in caesar model, how to enter desity of the fluid for trunnion, is it 0 and ambient temperature should be considered? help me wih a example.

Step 1 :Model a rigid element from center of the pipe to the Surface of the pipe with Line Propeties.
Step 2: Model a trunnion with ambient temperature and without Fluid density and pressure.
Step 3: Apply SIF as per the attachment.

Hello Chander,

Please learn to use the search function. Look for the "search" window above the calendar to the right of this page. This is a topic that has been greatly discussed here and there is much you (and I) can learn from reading all the past discussions on this.

Respectfully, John.

You should not apply SIFs here. This is not a "tee" or other "Code defined" fitting. You did not cut a hole in the pipe and affect the pressure boundary. Therefore, Code computed SIFs are not applicable here.

Does the trunnion affect the local stresses at the point of attachment? Sure, but the "Code computed" SIFs don't address this situation.

Mr kellog?

Shear lugs and raiser clamp.

Thank you Ravanan,


Regards,

Chander

Many consultancies use code’s “reinforced tee” SIF for trunnions on straight pipe. The reasoning is like this…

There is no weakening hole cut on the pipe for a trunnion. The pad used in a reinforced branch compensates for this removed area and we get a branch like situation. So in effect a trunnion on straight pipe without pad is equivalent to a reinforced branch for SIF computations.

Of course this is not accurate but a reasonable approximation in most of the cases…

The method of using a "branch" in the model addresses some but not all the possible stresses that "hollow" welded circular attachment will create on the pipe wall.

In & Out Plane bending will be reviewed but punching or axial loads will not be addressed, neither will torsional shear (depending on piping code)

So as you see its an approximation just as in all things....

In case of high axial loads we use pads...again the SIF is only a reasonable approximation!

I don't like pads on hot lines especially if there is any kind of tension load. See image

This is a question in response to R. Ay's discussion on trunnions!! Is there any consideration for stress intensification factors involving trunnions based on non-uniform geometry where nominal stress equations apply?? For instance in FE/Pipe, there is a template set up just for base ell/trunnion supports of which, in the output, calculates inplane and outplane SIF's for the trunnion and bend. FE/Pipe claims that these SIF's can be used in a beam-type analysis such as Caesar. Also, how do we account for the absence of flexibility in the elbow with a trunntion attached?? I know these are rookie questions but alot of people here in the office have been asking these same type of questions.

use the search function first....

What I meant in my post above was that you should not apply SIF values obtained from using the "standard Code equations", found in Appendix D of B31.1 and B31.3. These equations do not address trunnions.

If you're using FE/Pipe, then yes you can take the FEA generated SIFs and apply them to your CAESAR II model. As to the reduced flexibility of the elbow, if FE/Pipe gives you a value for "k", then you can use this as input to CAESAR II also.

Note, these values and assumptions must be properly documented, since you have just stepped outside the scope of the Code.

thanks for the quick reply....Mr. Ay you still prove to be inspiration to me because I sound so smart telling my boss what you said in my own words....lol!!! Just kiddin....I am just a little college kid trying to get ahead in the world!! Thanks for clarifying that for me and have a good week.


p.s.

Mr. Luf I did use the search function first...that is how i found this topic!!!

A few thoughts.

WRT re-pads,

On an Elbow, i sometimes do not ask for a re-pad but try to bump up the (elbow) schedule, as the re-pads are hard to form.

On straight runs, same again with the pipe.

Although doing the above, i do not indicate a pad in my calculations, but show the increase in wall thickness.

WRT the photo

* That looks like bad welding to me, the break is far to clean.
*The pad looks too small and may well not be helping to spread the load.
* The pad is not circular and may be accentuating stress concentrations (De- Havalland Comet windows...)
* The Pin in the joint looks ok
*Pad not insulated - differential thermal growth?
*For the size of the pin joint on the bellows ( and that is some bellows) the attachments LOOKS way to small.
* The bellows looks old, cyclic fatigue could be a factor.

Having touched upon the topic of re-pads, I have a question on re-pad types. Pl. see the attachment. In what cases are type 1 (rectangular full pad) and 2 (oval hollow pad) used?

The first impression is that type 1 is preferable for heavy axial loads and type 2 for heavy bending loads (application for supporting vertical pipes). But we did an comparative FEA and the results showed for all types of loading type 1 is better…so I’m not sure if there any other factors involved in their respective application….

SamAbdul,

Your type 1 is just a doubler or wear pad, normally used beneath a resting line to better distribute concentrated loads.

Trunnion type 2 ... If the load is mostly axial along the trunnion, a pad without a hole is ok, functioning much like 1 above, since you are then dealing mainly with punching shear at the pipe wall / trunnion interface. However if there are significant bending loads, the same welding philosophy as a ring reinforced tee, i.e. a fully welded ring is far superior. To explain this, consider the same connection but with the trunnion welded only to the pad, which is then welded to the pipe. The tensile side of a bending load will tend to pull the pad away from the pipe. This will bend the pad causing a tensile stress at the the back of what is an uncontrolled fillet weld. That is potentially a massive stress concentration, dependent upon the pad proportions. If cyclically loaded, rapid weld failure may occur.

So IMHO, stick to your existing type 2 construction detail.

Moverz,

I agree with ur views. As I indicated earlier this was my first impression as well. But the FEA results had me confused. The presence or absence of material just below the trunnion had a big impact in the results, for all types of loading. I did not do the FEA myslef (our Abaqus FE spcialist did) and I'm not sure how he accounts for the weld boundaries?

WRT " which pad is what"

I suggest reading a good pipe support book before using expensive FEA.
A good browse through the Kelloggs manual will do wonders.
( and if anybody knows of a good book then Tell us all.)

I don't pretent to fully understand all the minutia of what goes on in a re-pad/pipe intersection, but i have a good idea of the basics and try to
apply common sense to any of the applications.

Hence i resist the urge to use wear pads as re-pads.......

Superpiper,

Even in some of the world’s best EPC companies’ standard support collections, the type 2 pad option i.e. re-pad is not available. The wear pad (type 1) is being predominantly used as a pad for trunnion.

chinese whispers syndrome for pipes!

Just because a support appears in a standard, this doesn't negate the reader from thinking.

Our pipe supports are shocking and i refuse to sign isometrics which show them.
In fact i only approve support location and function, such is the state of our standards(work in progress)

I wouldnt use a wear plate as a repad.
The fact that such was used in the afore mentioned picture was probably the reason it came off.

From eyeballing that picture, the plate looks totally wrong.

If it looks right, it probably is right, and if it looks wrong, it probably is wrong.--- never a truer word said.

An interesting historical note... M.W. Kelogg company people who explored this issue before most of the participants and readers of this forum were around specifically excluded "Repads" from their analysis and design standards. Why ??? beats me!

However putting a heavier pup piece of pipe with tapered butt weld ends into thinner wall pipe is usually a fairly easy thing to do. Also a big issue to be considered is just how attached is the repad to header????.... most calculations assume the pad and header acts as one solid piece but thay aren't.

The failed pad in the picture is the worst I have come accross. As I mentioned on hot especially creep range lines any attachment weld, pad and similar tend to fail at the end. It is only a question of time. Unfortunately I was not on site when the failure happened so no close up inspection.
- small pad yes
- bad welding yes
- pad was insulated but it is possibly different material to pipe and in addition is cooler than pipe
- bracket is correct size. Pad should not be there at all. Trick for this design is inside the pipe.

This type of bracket without the pad will fail also if temperature is higher. It will actualy fail even if there is no external force. Very high temp designs have floating rings with shear lugs for this reason. Basically same design as raiser clamp.

You weld a trunion on main steam line on utility boiler without attaching anything to it and after 20 years there is a crack. We have found cracks even in gamma bosses. (Plugged and reinforced small hole to get radioactive source inside the pipe)

Easier and safer to use raiser clamps. No FEA, no problems in getting special pipe, minimal welding and minimal failures. I just checked 22 year old 800NB steam line which had really bad hammer. No damages what so ever on clamp supports with lugs. Next project is P91 MS line where somebody put trunions only on one side of a vertical pipe. Will be expensive for the plant owner to fix them.

Cold lines are different.

Superpiper,

IMHO questions arise only out of thinking!

If something is specifically excluded in a company's standards collection, there shud be some solid rationale behind it. I'm just trying to find what cud b the basis?

And what is the point in approving support type and function, if something else is actually implemented & constructed? It is sheer escapism!!

Jouko... another old guy by the name of Ron Haupt (Not as old as John Breen mind you) once stated "Trunions" do not belong on steam systems in power plants.....

My own opinion is that once you go into the creep range I try to minimize welded load bearing attachments as much as possible.... and I take into account the creep weld joint strength reduction factor as found in B31.3-2004. It will derate the strength of the weldment significantly above 950F.

B31 main has an ad hoc committee putting together common wording to be put into both B31.1 and B31.3 books concerning "W" B31.1 has in their own "wisdom" been unable to agree on what to do concerning "W" despite the fact the failure of a weld seam that killed 10 people that stimulated research in the creep weld strength reduction problem occurred in power plant!

Committees require consensus and sometimes one or two people with strongly dissenting opinion's make things difficult.

Returning to the many references to company standard pipe supports etc.

Beware of slavish adherence to these things. Company standards even in the better operations can be severely limited, particularly if an off the shelf supports software package has been attached to a 3D CAD design system.

Be prepared to design a decent special support. If it's good, it will be in the 'Standards' pretty soon.

A picture of that riser clamp would be intresting......

Very good points from Jouko.

Unfortunately I did not take a picture of the supports on site. You can see raiser clamps for instance from Lisega cataloques. Download the set from www.lisega.co.uk. I do not have addresses for other suppliers on hand just now. Maybe some people can help here.

Image is for Lisega new type. Traditional design is based on formed clamp, which need often oven, heat treatment and heavy press. This is like puzzle made out of plates. Cheap and easy to make in large quantities. For what ever reason I actually "checked" their design when they came on market. There is a patent as far as I know.

I modelled today MS, HR, CR and bypasses of a system done some 15 years ago. Designer new only trunion. They stick out to all possible directions, even downwards. This %^&*%$# put even dummy branch/trunion on bottom of a pipe set on a cross beam hanged by two constants above - pipes in creep range. Then there is 17 000 lb valve in thin air. Absolutely the worst design I have ever seen. In 2 months time we will know how much of the pipe is damaged.

On this topic, are there equations (hand) I can use to calculate SIF's and flexibilty factors for a trunnion supported elbow? And if so, is it valid to apply these in my caesarII model? Right now I don't have access to FEA software.

I did find an interesting article by Dave Diehl in which he suggests an equation, but I'm not sure if it is applicable, and my situation does not pass the requirement: T ≤ t ≤ 2T
article is here:
http://www.coade.com/newsletters/oct02.pdf

Any suggestion of reference material or where I can look for answers would be greatly appreciated! I have read a ton of material and discussions (online) regarding this subject, but I still feel a bit lost and would appreciate some feedback from more experienced engineers.

G. Archibald,

Dave Diehl suggests just flexibility factor k. Not SIF.
SIF please look at these articles:

Stress indices for straight pipe with trunnion attachments

http://www.epriweb.com/public/TR-110162.pdf

Stress Indices for Elbows with Trunnion Attachments

http://www.epriweb.com/public/TR-107453.pdf

Thanks Mav!

Dear all friends,

Sometimes our knowledge and understanding are limited to specific software and some specific codes. Why don't you use other software to handle some specific calculations without more headache? Or you may be in doubt at all.
Many critical lines must be evaluated by both CAESAR II & FE/Pipe, specially for intersections, trunnions, stancions, shoe welded to large pipe, etc. For accurate and reliable solution for intersections you need to consider both SIF & flexibility in all directions. In FE/Pipe use Unreinforced Fabricated Tee template and set No Hole in header in order to model trunnion and calculate SIF & Flex. Attached to this post, there's an useful document that I've found in PAULIN website. I hope it helps.

Regards,
Farhad Salehi

I prefer to use experemental data for SIF than FEM results because if we use shell elements FEM couldn't calculate REAL peak stress. To calculate real peak stress we need to use volume FEM elements with very fine mesh at stress concentration zone. FE/Pipe can't do that. There's just special factor that user should enter. Result FEM stress multiplied by this factor to calculate REAL peak stress. Where should we take that factor? It could be obtained only from experiment...
Sometimes FEM analysis give us qualitative incorrect results. For example for 45 degree branch...

This is not the method of welding trunions to pipe with a repad. The trunion pipe is welded directly to the main pipe with a full penetration weld. Then the repad comes on and is welded to the main pipe on the outside of the repad, and then the inner seam is welded to the trunion pipe. That way the repad only adds strength, it does not take away strength. These are my thoughts, please correct me if I am wrong. The picture you posted shows the linkage welded to the pad, and the pad welded to the pipe. Not as strong...

I agree with SAM's argument as SIF is nothing but a factor to increase bending stresses. Any attachment on a piping will result in discontinuity stresses ie local bending stresses as per shell theory,although such stresses wll die down as functions of SQRT(RT). Hence,in lieu of a more sophisticated analysis, a factor to account for additional bending stresses is something better than nothing.

Based on some comparisons that I did between FEA and Kellog method for Trunnion attachments, I have noticed that the conventional Kellog method is quite accurate in computing the local stresses( this is based on my testings only and not a published paper). Kellog method is based on a cylinder loaded with a ring of uniform bending moment.I think in lieu of SIF, if local stresses checked , it is also an acceptable method.

Regards

Further to my previous post, I feel that to the extent an analyst has to go in for ( for example,properly computing peak stresses in FEM , incorporation of ovalization effect of supports etc) is something, an analyst has to decide based on the system in hand , his/her experience and his/her level of expertise.

For not so critical systems ( something cannot be quantified but to be assessed by an analyst), use of SIF or use of Kellog type of analysis is acceptable in my opinion.

Regarding FEPipe, my opinion is it is too tailor made with lot of limitations and can never be at par with FEA codes like ANSYS or ABAQUS and I dislike what it says " FEA MADE EASY".FEA should not be made easy as wrong results can easlily escape an analyst w/o proper training on FEA.

Regards