Specifically looking for guidance from Coade.
(Other opinions welcome)

Problem:
Emergency relief vent
Glass Lined Pipe
Runaway reaction.

Synopsis.

Glass lined reactor has glass lined pipework to the Condenser and then onto the Bursting disk.
After Bursting disk, pipework changes to SS and runs to relief tank.
Runaway reaction in reactor causes pressure to rise, causing burst disk rupture at 2barg
Pressure continues to rise until 40 secs later it peeks at 8barg.
There are 2 bends before the BD and 4 bends after


Problems

I have two main problems,

1.Relief forces.
I have two relief force scenarios here, i intend to do:
a) one run with the 2bar dynamic forces acting on the first elbow before the BD and on each elbow after the BD
b) The pressure rise to 8barg after the Disk has gone is over 40sec and can be ignored


2.Glass lined Pipe.

The vendor has provided nozzle charts and has asked that the pipe flange forces do not exceed same. Pipe Stresses are to be kept below b31.3 Alllowables to avoid yielding.

I have had no account of pipe deformation allowables from the vendor and can only think that
either to some extent, the glass is ductile, or that the vendor is mistaken?

To analise the forces i have placed a cnode at each flange, i have also split each spool into two to asses midspan forces and moments.

So, Am i going in the right direction????

Regards

Well timing is all....

Just got to analyze some SS piping on a glass lined tank.....

Imposed 5KSI bending limit and advised the field that they should contact the vendor... heres the vendors response...

According to the vendor, tank nozzle load limits are 100 lb-f. per inch of nozzle diameter, which means 400 lb-f. for our 4" nozzle. According to the vendor, the nozzles are only designed to support the downward force (weight) of an isolation valve and any additional weight must be independently supported. The 5KSI bending limit meant nothing to the vendor rep. I spoke to. How does this figure into your calculation?

So there you go questionable advice from a vendor with an intermediary involved who is also unknowledgeable.... perfect!

Don't all shout at one chaps...................

Thanks anyhow John.
Pfaudler give for 4" nozzle 0n STD wall CS nozzle with glass lining:

FZ(Axial)= 3500N
FX=FY = 750N
MX=MY = 400Nm
MZ = 250Nm
It is my opinion that these are conservative.

For Glass lined pipe, they have confirmed that upto yield, the glass lining will take deflection, including the 1% ovalisation as allowed by ASME


For your problem, do the best you can to get the figures as low as poss, then pass onto vender for approval. Usually works.

Bit disapointed with the Coade response.
'Don't Know' would have been appreciated at the very least.

Ok - "I don't know".

Superpiper,

Thanks for your comments…

I am waiting for the so-called designers (so-called because simple connecting the dots in space is not really all there is to designing a piping system) to come back to me as to what steps they want me to take to decrease the over strains in the system.

Ultimately I had planned to submit the final loads to the vendor for review. (Once they sign off on the loads the design is truly done)

As for your disappointment with COADE I suggest you may be expecting too much from the wrong party.

COADE produces analysis software that must comply with many codes and provide useful features. As far as glass lined anything is concerned they do not manufacture said products. Therefore it is out of their scope of responsibility and knowledge. The people who manufacture and sell this sort of stuff should not only manufacture it but should tell people how to use it. This information I would liken to an owners manual etc. But it seems as though some of these manufacturers do a poor job of handling THEIR responsibility.

Yep. i know. I was rather hoping for a critique of the (BD)technique.

Still. good reply by richard.

TJN,
The thought "to some extent, the glass is ductile," brings to mind that there are glass piping systems having some design basis. As memory serves me, glass is elastic in behavior until it fails - essentially no plastic yielding occurs. So 'ductile' and 'glass' are a paradox.
An old SCHOTT Process Plants and Piping handbook dated 1979 had data for their 'Duran Jena Glas'.
Modulus = 63 kN/mm2 (9.13 E6 psi)
Tensile bending = 4 N/mm2 (580 psi)
Thermal expansion = 32 x E-07 , alpha 3.25 E-06/K
or (0.0022"/ft/100deg F)
Steel would expand at a rate 3.6 times the glass, however the glass modulus is less so the glass lining could accept reasonable temperatures.
The real limit is the 4 N/mm2 tensile. It is an order of magnitude less than the 5 ksi limit imposed by John Luf, but that 5 ksi includes the strength of the steel.
It might be advisable for the owner to inspect and be ready to replace some of the glass lined piping whenever the BD is replaced after an event.

Perhaps a good approach would be to assume that the glass lining carries no load, but is rigidly fixed to the carrier steel. Then working with strains, we could estimate what load the nozzle could withstand, before the glass reached it's tensile limit [4MPa seems about right from the info I have].
Quick rough estimates I've done suggest that the figures Pflauder are giving [for the axial at least] are probably reasonable. As simplified as they were, they suggest the glass will fracture under axial load [for a 4" STD nozzle] of 28kN, which suggests a factor of safety of around eight. I've not estimated any other loads.
You may need to consider including some sort of dynamic amplification factor for when the BD goes off.

What if you try the line as a jacketed pipe, and in the inner pipe you input the properties of the glass? Do you have those properties?
I woul say that glass doens't have any dutility, is fragile, and if can be compared somehow with any metal would be one with SU and SY very, very close, and the elastic modulus will be a lot higher than steel. And in top of that is very dependent with temperature, in fact, you could blow and make a bottle!
Finally vendor loads are vendor loads, It is very hard to bite the bullet and override vendor limits (even when you prove they are wrong), and even harder if you don't have any code backing you up.

Regards,

If you don't mind to tell us how your story ended.