anybody else here find the resolution of bellows
in CAESAR II hard work.?
I have numerous bellows, both untied and tied in the same system.
Add this to glass lined pipe which have low flange force allowables, and i get giddy.

I wish it was easier.
Every time i switch on caesar i change my mind about where and how i am goint to apply pressure thrust forces which are VERY important in my glass lined environment.

(my analysis at the moment includes:)
Glass lined reactor
Untied bellows
Tied bellows
Glass lined pipe
Bursting disk
Anchors
Graphite block exchangers
Rotajoints???
Stainless steel vortex tanks
45deg elbows in the relief line.
Fire cases
Cryogenic.
Vacuum
Explosive runaway reactions
Oh and i am on my own, nobody to bounce ideas with

Im going for a lie down........................

I would love to be involved in a cutting edge problem as you are now. Because I'm feed up with simple problems... There is one thing that you should keep in mind: Solve one problem at a time.
This mean try to break your huge problem in smaller ones, solve each problem (or get a plan of action about it)and move to the next problem. Don't loose the "big picture" because at the very end the sum of al solutions probably wont be the solution, but a real close one.

Regards, and don't give up

I have found the absolute best most foolproof way to use bellows is as a hinged joint this negates pressure thrust and yet still allows a low force rotation to occur in the piping system... I love hinged joints!

Where to locate static pressure thrust around expansion joints?

CAESAR II will add thrust loads on either side of the expansion joint if you provide effective ID. This is not truly the right place but its convenient. And I really don't see a difference between pushing a pipe and pulling a pipe.

If you want to place it correctly, I suggest this...

Imagine you are standing inside the pipe and looking upstream and downstream. The surfaces you see are where the thrust loads for the run you are standing in are applied. The area in your (P times A) calculation is the projected area in a plane perpendicular to your line of sight (the line along the pipe axis). This area includes the last convolution of the bellows, any reducer and, of course, the back wall of the bend. Around equipment, the thrust load is applied to the back of the equipment and that load is usually absorbed by the equipment "anchor".

For glass lined pipe where Flange forces are everthing:

Vessel-Bellows-Anchor-Bursting disk-Elbow

Pressure thrust is added before anchor, not afterwards, putting a positive force on the anchor, and compressing the flanges before the anchor, instead of putting the flanges in tension after the anchor, which is critical when adding a Burst Disk relief force onto the elbow.

Why is this so important?

Glass lined pipe manufactures give a requirement that flange forces should not exceed xxx.
However these forces are for a rating (in my case)
of 10bar@200c
This 10 Bar is stealing all of the allowable pressure thrust (and hence) axial force that is allowed for by the vendor.
My system only goes up to 2bar at max.
If i add the pressure thrust at ALL to the elbow, then the flange forces fail.

My solution has been to ignore pressure thrust,
Give the vendor the flange forces and moments as calculated by Caesar from thermals only,
and include an iso, indicating where in the system, the pressure thrust will result.
I am hoping that i can use the difference between
the 10bar rating and the 2bar actual to enable the thrust force from the Bursting disk event
In effect i am moving the resultants from the pressure thrust around the system, the location of which is critical.

To simplify, if i do as Dave suggests and apply 24000N of force on my elbow, plus the recoil force, then my vendor will hurt himself laughing :p

seems reasonable to me, but then on the other hand its not my glass lined stuff....

Glass = Bier Glass =

I've finished, and for 'Massabie' here is the story:

Glass Pipe:
The vendor originally indicated his nozzle loading chart should be used to reduce the forces on the flanges. this reduced the Sa to a unworkable level. I created a force matric for the more overloaded flanges and sent them to the vendor to have the ensuing stresses accepted.
I removed the pressure thrust from the matrix and informed the vendor of the working pressures, hoping that the vendor would use the minimum amount of pressure stress in his calculations.
Upon reciept, the vendor contacted me and i had a very disjointed discussion which had me worying as to how the vendors would asses the stresses
A few days later, the vendor returned to indicate that i should not be treating the pipe flanges as nozzles (REALLY!?!?) and that:
1.Glass lining is as strong as the steel pipe.
2.Glass lining is flexible.
3.Glass lining will not yield and therefore the stresses in the piping should be kept to within Sh as per appendix A b31.3

This in one sweep reduced all the stress to an acceptable level

The bellows have been shoved in the design by the designer without any thought as to what they do, I have had to create data sheets for the vendors to ensure that the tie bars are correctly configured so as not to tear themselves to pieces

I extracted myself from all responcibility for the Rotajoints. I had the vendor asses the resulting loads and state in writing that they were fit for purpose in a relief system.
What happens when the axial stiffness is removed in a pipe during a relief event is beyond my limited experiance. and i let the client know this.

I did a static equivelent for the dynamic forces during the relief.
The configuration is quite simple and the forces are low. F=pa*DLF(2) was used for the forces on the bursting disk

Although exposed, the pipe runs are small and are well obscured from the elements, hence wind was discounted in the analysis.

Supports
The systems were anchored just after the tee for the vessel and the condenser. this removed any torque from the bellows which resulted from thermals. The anchor was a 6" xs trunnion (4" long )welded onto the back of the tee.
This allowed the tee to be a thicker fitting (xs instead of std) and meant that the glassing process would be simplified (a tee instead of 6m pipe)
Whilst not strictly correct, i summated the stresses as calculated by the kellogs method for trunnions and kept these bellow Sh.
A 6" trunnion is a strong thing and i would be suprised indead if the local wall stresses could be a problem in the tee
For the record, GL pipes should be supports from pre-welded clips.
Clamped shoes are not recommended.

I'm still unhappy about the addition of the pressure thrust forces in the caesar calc reports and i have marked them up as appropriate.
next time i will probably add the forces myself


Thats it really, just to say that my work is due to be checked by an outside consultant.
I'm sure if you were to look hard enough, errors will be evident, but overall i am happy that the system is acceptable for service.

"The bellows have been shoved in the design by the designer without any thought as to what they do,... "

Hard to believe.... yeah right see this all the time for over 35 years....

Thanks for the update. I still have problems with glass being flexible (just not "intuitively obvious
I guess)... but I haven't ever looked into that,
so maybe it is...

-Shannon

SUPERPIPER,

Be sure the let us know who this glass vendor is. I'm definitely going to pull for the strong as steel glass vendor the next time I have to suffer through a glass lined system.

Also, with regard to your pressure thrust loads on the glass lined flanges, I would expect there to be a difference in allowables from compressive forces (which would tend to want to crush the glass lining on the flange face) vs. pulling forces (that would tend to want to create leakage). Since the pressure thrust would tend to want to open up your flange pairs, I would think you be able to work with a higher allowable value.

Pfaudler-Balfour in Scotland.

To be honest, i doubt they could be as precise as to give different values on the flange face.
(another example of how carefully one should asses the thrust loads) But i never thought of that angle. good thinking.

But i will say this, the glass is not glass that you see in windows, and the idea that you do not yield the base material, seems perfectly logical.
Apparently, the engineers in Scotland have seen test strips of this glass being bent nearly 90deg.