Spring in Regulator Question 1

[mcardle27]

Hello All,

I am working on a problem with a diaphragm operated regulator where the regulated pressure has been dropping over time. The regulated pressure is set with the main compression spring which sits on top of the diaphragm and is compressed using a set screw. I have done analysis to show that the spring is over stressed at the target regulated pressure. I am assuming that the spring, while sitting in the regulator, is losing force, and thus not as much pressure is required to balance out the spring force.

My question is, why isn't the spring gaining spring rate, or becoming stronger? Couldn't this initial compression and yielding almost act as an set removal operation that would strengthen the spring?

Our initial thought was that the spring will be over stressed and thus yield, but since its motion is restricted on top by the screw, it will begin to lose spring rate and thus the regulated pressure would drop. I don't know if this is correct now. Maybe I am thinking about this the wrong way, but someone suggested that the spring would become stronger and now it has become a nagging thought that I can't seem to think through. Any help would be appreciated.

 

[TheTick]

If the spring is yielding, it probably yields just once to match its overstressed form. Additional cycling will not make it yield again.

 

[mcardle27]

Thanks for the reply. I actually was able to perform testing that supported this too. I guess I am more concerned with the initial yield though. Does is make sense that initial yielding would result in a lower regulated pressure? I was concerned that the initial yielding would make the spring stronger and thus the regulated pressure would increase.

 

[israelkk]

A spring is going through a relaxation process where it is compressed to constant length and looses force (the rate stays constant). When a spring is compressed under constant load it is shortens due to a creep process.

In your case the spring looses force probably due to relaxation or over stress. The larger the stress in the spring the larger the relaxation rate even if the spring is not over stressed.

Rate is defined by the active number of coils, wire diameter and outside diameter. Therefore, the rate can not increase unless some coils are come to touch each other during compression thereby gradually reducing the number of active coils. However, the spring should be designed this way and will have a non linear force-deflection behavior.

In your case, even if you do not yield the spring it will relax with time. How much depends on the shear stress in the spring compared to the ultimate shear stress of the spring wire.

http://israelkk.googlepages.com/home

 

[zeusfaber]

The thing to remember with spring regulators is that it isn't the spring rate that determines their outlet pressure - it's the preload on the spring when the valve is just shut.

If the spring yields (or just relaxes) at target outlet pressure, that reduces its length. The force on the spring cannot increase while this is happening (the valve stays shut, so with nothing to drive a force increase, the force must either stay roughly the same, or reduce). That means that for the spring to return to its original length (the length needed to open the valve) the spring needs to see a smaller force - hence reduced outlet pressure.

A.

 

[mcardle27]

Thanks for all the feedback.

zeusfaber --- This helps make some sense of it.

Could I look at is almost like a permanent deformation on the stress-strain curve. Where the spring is pushed past is elastic point, and now it rides a lower sloped curve on its way back? Thus for an equal amount of strain the force output will be less?

I also get that the spring is relaxing (due to over/high stress). Since the spring is essentially at a fixed length, it losses force and hence outlet pressure. But I am struggling with the fact that set removal techniques are used to strengthen the spring? I am probably just having trouble thinking through it.

 

[israelkk]

mcardle27

"set removal", "preset" or "scargging" as it sometimes called is a secondary operation that the spring is gone during the spring manufacture. But, the spring must be designed for that process to take advantage of it. It is more costly and is being used when space for the spring is limited (mostly due to an error in the preliminary system design that didn't left enough space for the spring and left the spring design to the end of the process). Preset spring design is not for the casual spring designer, you need to be a very experienced spring designer expert to know how much preset you can introduce into the spring wire to fully utilize the spring wire.

Without the exact details of the spring dimension, material and load vs deflection data, etc., there is no way to tell if a preset designed spring can be utilized in your regulator.

http://israelkk.googlepages.com/home

 

[tbuelna]

The travel of the relief valve will be due to a combination of the opposing forces created by the diaphragm and the compression spring. Compressing a helical compression spring beyond the point at which the wire yields creates a "compression set" in the spring. This reduces its free length and will result in a reduced force at a given installed length, but should not usually have much effect on its linear spring rate. Since you have an adjusting screw you should be able to compensate for the loss of preload in the spring due to compression set from a given amount of valve travel. But if the spring is subject to repeated compression set after re-adjustment, then you will need to modify your spring design.

 

[israelkk]

tbuelna

It is always a good design approach to design a spring that will not yield even when the spring is compressed to solid. In fact this is a requirement for all US navy springs to be designed that even in solid length there will be no set.

As to your statement "Since you have an adjusting screw you should be able to compensate for the loss of preload in the spring due to compression set from a given amount of valve travel", this may be true only if the spring after "setting" retains enough spring deflection till solid to create the desired spring force.

 

[Tmoose]

If this is a single stage regulator mounted on a compressed gas cylinder I would expect the pressure to drift downward slightly as the cylinder pressure drops when the cylinder empties over time.

I'd measure the spring free length now, and check it again if the pressure drift occurs again.

 

[zeusfaber]

Word of caution about that: Lots of spring regulators (especially those with an upstream valve) answer reducing inlet pressure with a rise in outlet.

A.

 

[Lou Scannon]

Word of caution about that: Lots of spring regulators (especially those with an upstream valve) answer reducing inlet pressure with a rise in outlet.

This is true but the third dimension to outlet pressure function is the volume flow rate, which, as it increases, will eventually trump negative droop.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz