Limiting Stress during test pressure 1

[soletto]

Hallo guys,

i was asked why i do limit the allowable stress while test pressure to 90 % of yield stress at test temperature.

Now i know that it is common praxis to use 90 % as additional safety factor. But is there a clause in the Norm anywhere so i can add this link?

I cant find this 90 % limitation.

Thank you br Stefan

 

[LittleInch]

Any particular design code you're thinking of?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TGS4]

There is no such Code limit in ASME Section VIII, Division 1. There exists a 95% yield limit in ASME Section VIII, Division 2. And a 100% of yield limit in ASME B31.3.

I would note that the limit applies to the primary general membrane stress - the one that you would calculate with hand-calcs. Obviously, the membrane stresses at discontinuities (such as nozzles, head-to-shell junctions, etc) may exceed that value during the pressure test. And the membrane-plus-bending stress (range) can go up to twice yield under normal operating conditions, so there's that.

 

[SnTMan]

I don't know of any original source, but any number of client PV and HX specs impose this limit.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[saplanti]

They might be forced to have additional safety factors by their insurance companies. Sometimes they have statistics on failure locations, so they may ask a new factor in the design to cover up some unknowns.

These are not related to code rules. If we look at codes they are the minimum requirements, and clients can ask more requirements that will not against the rules.

 

[LittleInch]

It's because you're testing the item not trying to break it.

Yield stress involves a level of permanent strain as it is defined commonly as where the slope of stress vs strain becomes half the elastic slope.

So 90% is the highest practical level to go to without risk of permanent strain.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

See ASME VIII Div 1 UG-101 PROOF TESTS TO ESTABLISH MAXIMUM ALLOWABLE WORKING PRESSURE.

Regards

 

[Vikko]

In EN 13445-3 tab. 6-1 the 'f_test' nominal design stress for test or nominal test stress (in ASME BPVC it is called 'S' maximum allowable stress) to be considered for the test condition is roughly 1/1.05=95% of the yield strength at ambient (or test) temperature (except for bolts, see 11.4.3.1 for that).

 

[r6155]

@soletto
Perhaps you heard something similar to 90% yield stress in bolts as stated in ASME VIII Div.1 UG-99 (b).

"Bolting shall not be included in the determination of the LSR, except when 1.3times the LSR multiplied by the allowable stress of the bolt at its design temperature exceeds 90% of the bolt material specified minimum yield strength at the test
temperature".

Regards

 

[IdanPV]

ASME Section VIII-1 not fully address this issue.
I think that the limit of 90% of yield, which a lot of designers use, came from ASME Section I PG-99.1.

Another option which suggested by TGS4 is to use the limits as per ASME Section VIII-2, 4.1.6.2.
Which allows you 95% of yield.
This is appliacble when using ASME Section VIII-1, U-2(g) in conjection with Mandatory Appendix 46.

Since the code does not adress all aspect of design, my opinion is that the 90% yield limit is a good engineering judgment.

 

[r6155]

@ soletto
How do you plan to control that stress does not exceed its value in each pressure part during the pressure test?

Regards

 

[soletto]

its for the analytical calculation and be code compliance. I just heard that 90 % is common praxis but i didnt found a source in the Code.

Test Pressure is beeing verified by analytical code calculation for all pressure containing components like branches etc.

We use 90 % because of common praxis thank you i will look into the code you wrote.

But the crazy high test pressure in B31.3 pT=1,5*pD*allowabe Sress Ratio is really crazy high....

Verifying via flexibility calculation

 

[SnTMan]

soletto, I believe the word you are looking for is "practice".

Hydrotest multiplier for Sec VIII, Div 1 was 1.5 x stress ratio for many years before allowable were increased (in the '90s, I think), when it was decreased to 1.3x.

Much replacement vessel and parts work is still tested at 1.5x.

The 90% yield limit in client specs goes way back as well, again I don't know of any original source.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[TGS4]

The B31.3 test pressure is not, in fact, crazy high. And limit in B31.3 is 100% of yield.

And to reiterate, there is NO limit on the stresses during the pressure test in ASME Section VIII, Division 1.

 

[r6155]

I disagree
No clarification is necessary because it is obvious, the limit is indicated in the material specification. Some users set a limit of 90%, 95% of Yield, etc, etc ......, but never more than 100%.

Regards

 

[Some Curious Guy]

Usually people ask why the hydrotest factor in Sec VIII Div 1 is 1.3 and Div 2 is 1.425. To that we would answer that the testing at 1.3 MAWP would reach 90% membrane stress yield limit and 1.425 MAWP would reach 95% yield limit.

For Div 2 the limit 95% of yield was given in code so people back calculated and knew why 1.425 hydrotest factor was applied.

For Div 1 the 90 % yield limit was not given in code. Hence Its my guess that people calculated how much stress would 1.3 hydrotest factor result in and they arrived at a figure of 90% yield. Its just a guess though.

I do not know which code you are using. Just take care the Stress limit ( 90% Sy, 95% Sy or 100% Sy ) is in some agreement with hydrotest factor ( 1.3 or 1.425 or 1.5 )

 

[soletto]

Ok thank you,

but whenn i consider a P355NH and calculate B31.3 test pressure with 1,5 times design pressure times the ratio of the allowable stresses i come to very high values for test pressure

But why then is in B31.3 the test pressure is calculated with the factor of 1,5 and the ratio of the allowable stresses.

Compared to other standarts where only the factor is 1,25

 

[Some Curious Guy]

I did not actually get your question. I do not use B31.3 much

I am assuming you are trying to ask why we multiply allowable stress ratio to mawp to find hydrotest pressure.

A general hydrotest formula is

Pt = factor x Pd x ratio of St/Sd

Pt is hydrotest pressure
Pd is design pressure or Mawp at design temperature.
St is allowable stress at test temperature
Sd is allowable stress at design pressure

Now Pd is the the value at design temperature while we would carry the hydrotest at room temperature or test temperature. We know that allowble stress of a material increases as the temperature goes down. So the value of design pressure at test temperature should increase in direct proportion to the increase in allowable stress. Hence we multiply desing pressure at design temperature with the ratio of allowable stress at test temperature to design temperature

Pd( test temp) / Pd( design temp ) = Allowable stress ( test temp )/Allowable stress ( design temperature)

So Pd ( test temp ) = Pd ( design temp) X St/Sd

Hope this answers your query

For the factors we can easily get them with some simple arithmetic.
In most cases the Allowable stress is 2/3rd of yeild stress.

Yeild stress is 3/2 of allowable stress

Hence for 90% yeild limit the factor would be
= 90% of 3/2 = 1.3

For 95 % yeild limit it would be
= 95 % of 3/2 = 1.425

For 100% yeild it would be
= 100 % of 3/2 = 1.5

Now I do not know how did people arrive at a factor of 1.25. Perhaps there is an another criteria for that. Hope other members can clarify that.

 

[soletto]

Hallo,

thank you for the post.

But:

B31.3 Requires a factor of 1,5!

so

Pt = 1,5 x Pd x ratio of St/Sd while St and Sd is allowable at test temperature and design temperature

that is higher compared to other standarts

AD2000 Requires the following:

1,25 factor is from AD2000

pT=max ( 1,25 x pD x SyT/SyD ; 1,43 x pD ) while Sy is yield at test and design

 

[r6155]

It is wrong to compare different codes. There are several elements that differ between them: inspections, destructive and NDE, WPS, etc.
For the same pressure vessel there are different costs depending on the code selected.

Regards