What is "Free length" of springs in safety relief vavles??

[kokerkov]

It sounds silly. But it has a great effect on our product.

Our springs cannot meet the requirement of “ASME BPVC 2023 Section XIII”, Part 3.2.5, which is :

(b) The permanent set of the spring shall not exceed 0.5% of the original free length. Permanent set is defined as the difference between the original free length and the free length after the spring has been preset at room temperature by compressing it to its solid height threetimes. Measurement shall be taken at least 10 min following the preset.

Where “original free length” is not defined, unlike ISO 4126-7：

The permanent set of the spring is defined as the change in the spring's free length as a result of a series of compression cycles to solid in accordance with the safety valve manufacturer's specification or other appropriate standard. The spring shall be compressed to solid at least three times before determining the initial free length. The spring shall then be compressed to solid at least three more times before measuring the final free length. The permanent set shall not exceed 0,5 % of the initial free length.

Our spring can manage to meet the ISO test, but our client said ASME test's "orignial free length" should be tested as springs are delivered. So anyone who are familiar with ASME can give us an explanation? Our best guess is from ASTM A125, which free length is defined as "length after test load is removed", and "test load" is "solid load if possible".

Thanks in advance!

 

[LittleInch]

I really can't see any difference here.

both say "original free length".

When you measure Original free length is surely up to you?

Basically they are just making sure you don't yield the spring when you compress it to solid / cant compress it any more....

 

[GregLocock]

Cynical spring designer says scrag (fully compress for several cycles) the springs at the supplier before they go in the box. Finding out how many cycles is necessary before length at n+3 cycles is only 0.5% would take a happy morning in the lab. Worst case scenario is that you may have to change the spring design so that it doesn't yield before going coilbound.

LittleInch - it is not unusual to design a spring either way, deliberately. For example with automotive road springs we design them to yield before coilbound so that we can achieve a certain length at a certain check load, to give consistent ride heights. At the suspension's full compression the load is taken by a jounce bumper, the spring is still not fully compressed.

 

[MintJulep]

If you go back to ASME BPVC.VIII.1-2017

UG-136 MINIMUM REQUIREMENTS FOR PRESSURE RELIEF VALVES (a) Mechanical Requirements (1) The design shall incorporate guiding arrangements necessary to ensure consistent operation and tightness. (2) The spring shall be designed so that the full lift spring compression shall be no greater than 80% of the nominal solid deflection. The permanent set of the spring (defined as the difference between the free height and height measured 10 min after the spring has been compressed solid three additional times after presetting at room temperature) shall not exceed 0.5% of the free height.

It looks to me like the unclear 2017 wording has been "improved" by a committee that didn't understand the original requirement, nor the English language, very well.

Regardless, "preset" or "presetting" contains the prefix "pre", meaning before.

free length after the spring has been preset at room temperature

That phrase means that you must preset the spring to find its free length.

I don't have the 2023 version, but 2017 includes a long list of "referenced documents", and ASTM A 125 is on the list; supporting your use of its (also vague) definition of free length.

 

[kokerkov]

I really can't see any difference here.

both say "original free length".

When you measure Original free length is surely up to you?

Basically they are just making sure you don't yield the spring when you compress it to solid / cant compress it any more....

Yes, it is different.

Let's say spring's free length is measured as delivered. Our springs will fail test. But as per ISO's method, we can pass. It is because spring's length will slowly extending after presets(yes, we did before delivering to client)

 

[kokerkov]

If you go back to ASME BPVC.VIII.1-2017


It looks to me like the unclear 2017 wording has been "improved" by a committee that didn't understand the original requirement, nor the English language, very well.

Regardless, "preset" or "presetting" contains the prefix "pre", meaning before.


That phrase means that you must preset the spring to find its free length.

I don't have the 2023 version, but 2017 includes a long list of "referenced documents", and ASTM A 125 is on the list; supporting your use of its (also vague) definition of free length.

Thanks for the info. That is useful. Is 2017 version still valid? I cannot find same UG-136 in 2023 version.

We did presets in our factory, but spring's length will be extending slowly after the preset. That's also why they said "measuring after 10 mins"

If we test as ISO's method, doing 3 times presets and measure the "free length", it will not have time to grow back, so our products can pass the 0.5% difference in length.

If "free length" means length as delivered, it will exceed 0.5% length differences.