Watertightness - spray testing: The physics of leaks

[MintJulep]

A common method of demonstrating that a thing is "watertight" is to spray water on it.

Typically, there is a (seemingly arbitrary) set of requirements for water pressure, water flow rate, water velocity (usually not specified where the velocity is to be determined), and sometimes nozzle distance from the thing under test. For example, MIL-STD-810.

But what determines the necessary values for these parameters?

Or, the actual point of my question - if a certain test facility can't simultaniously meet all of the required parameters, what compensation might be made to allow for an "equivalent" test.

Is a 20 minute test with a nozzle velocity of 100 ft/min and a flow fate of 100 gpm "equivalent" to a 10 minute test at 200 ft/min and 100 gpm?

What is it that actually challenges a seal and causes a leak?

My thoughts are that you need to consider the kinetic energy impinging on the surface under test, but I'm not certain that's sufficient.

It's easy to imagine a pressure washer causing a leak that a garden hose wouldn't. But it's also easy to imagine something that would never leak when subjected to a garden hose - no matter how long you sprayed it.

Your thoughts on the subject would be welcomed, and actual references even more so.

 

[TheTick]

Capillary action, perhaps? This would be somewhat time dependent.

 

[vc66]

Are we talking about keeping water out of a specific area, or a tube trying to hold water in under pressure?

By the way your phrased your question, I'm assuming keeping water out of an area. In that case, I agree it has to do with the kinetic energy imparted on the surface. I'd also imagine that there is some normalizing factor, so that "equivalent" tests may be performed.

V

 

[Unotec]

I'd have to ascribe to the kinetic energy, opposed to residence time.
Having said that, what is the core of your problem? A watertight tank/pipe should have different tests than a water tight pump volute, a coating or a cloth. On a tank you'd want the pressure of the fluid to act for quite some time to detect leaks.
But on a strictly spray scenario, I'd have to think that velocity is the determining factor unless there is a wetability component to the material being sprayed. If with time and as TheTick mentions, capillarity could play a role, then you have time to consider as well.


<<A good friend will bail you out of jail, but a true friend
will be sitting beside you saying ” Damn that was fun!” - Unknown>>

 

[MintJulep]

vc66: Yes, keeping water out of specific area - more specifically, a thing that may be exposed to rain, or wind-driven rain, or move under its own power through rain, or get washed on occasion.

 

[IRstuff]

Pressure, sometimes; dwell time, sometimes. There are lots of systems than "breathe" because of pressure changes caused by ambient temperature and sunlight. At night, when dew forms, the internal pressure relaxes and the system will suck in the dew with no water pressure whatsoever.

So, for a rigorous test, you need a pressure spray test as well as a static temperature cycle test.

TTFN

FAQ731-376

 

[btrueblood]

What drives a leak during spray testing = Dynamic pressure = 1/2 rho V^2. I'd say you could reasonably trade gpm and time, keeping velocity the same.

Bigger issue is what the customer would say if you propose to take exception to the spec.

 

[MintJulep]

And there exactly is the point. I am the customer.

 

[btrueblood]

Sorry, Mint, should've added that it's also implied in spray testing that any loose fitting covers, poorly designed splash guards, or poorly contained seals, will also be tested by the force of the spray. Again, the spray velocity dominates the force developed, but gpm has a say in the matter also, via the momentum equation.

 

[GregLocock]

When designing expensive electronic stuff to work underwater the hardest condition to seal is when it is floating on the surface, as the unequal pressure around the O rings (etc) stops them sealing properly, and the movement of the riing pumps water across the interface. Once the part is underwater everything works as planned.

As such I would be very wary about trading off between the variables in a spray test without some form of comparison test.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MintJulep]

Interesting.

Dynamic pressure X mass flow rate returns a unit of power.

Perhaps two test setups with equivalent power would be equivalent for an equal duration.

 

[MikeHalloran]

Presence of soap or salt in the water also makes a difference.



Mike Halloran
Pembroke Pines, FL, USA

 

[MintJulep]

Did the units wrong in my head last night. It's

Dynamic Pressure X Volumetric Flow Rate => Power

Does trading velocity and flow with equal power being considered an "equivalent" test make sense?

 

[vc66]

I would say that power was the normalizing factor I was thinking of in my earlier post.

Although, thinking about it practically--I don't know if they're equivalent. Something feels different to me between water at a higher velocity with a lower volumetric flow rate versus a lower velocity with a higher volumetric flow rate. Seems to me the higher velocity would do more damage. That's just my feeling though.

V

 

[Unotec]

I'd agree with VC66, however, the notion of a more prolonged exposure to the water will have more impact if capilarity has any part to play on the issue

<<A good friend will bail you out of jail, but a true friend
will be sitting beside you saying ” Damn that was fun!” - Unknown>>

 

[vc66]

I also feel as though they're doing different kinds of damage. The higher velocity (I'd assume) would be "cutting" (think water jet), while the higher volume almost seems like its beating the surface, as a heavy rain would.

Hmmm... Interesting problem.

V

 

[MintJulep]

I agree that there are probably different failure modes for different types of seals.

A lip seal would seem vulnerable to high kinetic energy - causing deflection of the seal and opening a gap.

A compressed gasket may be more sensitive to capillary action and the wetting properties of the fluid on the mating surfaces.

A labyrinth might be sensitive to both kinetic energy and individual droplet size.

Since my TUT has all of these seal types I can't ask for different tests for different seals - well I could, but I'd get flack for it, and at the moment couldn't justify it.

Also, my original test parameters are "historical" - which means nobody knows how or why they were selected. Most likely they were what some test facility that no longer exists was capable of doing.

It is an interesting problem however - might be a paper topic for some obscure conference...

 

[DesignerGuy16]

Look into the IP code testing standards. ie IP67, IP69K, etc.

I also just reciently sat through a Design News web seminar on sealing/venting electronics. Was pretty informative (although was 70% product pitching). The jist is that without vents a completely sealed product can go through a rigorous even if not intended change in atmosphere. So much that the pressure differences in the enclosure and outside of it will cause an almost vacuum like effect and actually suck in moisture past even the best seals.

The idea of a vent that would allow a vapor/moisture guard, but still be breathable to atmosphere would take almost all of the stress off of the seals. Allowing them to be much more efficient since they don't have to seal against differentials in pressure.

James Spisich
Design Engineer, CSWP

 

[HDS]

I can't answer your questions directly but be very careful of equivalents. There are always subtleties to the test. Even for tests that are not historical or empirical in nature.

As an example. There is a test for packaging called the ISTA test. One of the tests is to drop a weight from a certain height onto the center of a long box supported on both ends. A little bit of physics says you have a shorter drop with a bigger mass and have an equivalent test. This should give the same energy and the same results.

However, a box that will pass the original test will fail the equivalent because it can not support the larger mass that is resting on it after the drop.

In a similar way the water my drain off at a certain rate so that a different flow rate would create different levels of water a each location. A seal that is only wet in one test cloud be submerged by another.

(However this is still a very interesting physics problem)

 

[MintJulep]

And we have not even discussed droplet size, nozzle spray angles and the distance of the nozzles from the TUT - all of which will affect the velocity at the test surface, and many other parameters as well.