Flow Scan Transducer calls for no sealant
 

I'm installing a Flow Scan Fuel Flow transducer. It takes 1/4 " NPT. The instructions say that absolutely no thread sealant of any kind should be used.

I did not know that pipe thread could be liquid tight without sealant.

Am I wrong?

Michael-

Yes they are. Do not use sealant. Sealant can, and most likely will migrate over time and fowl the impeller. Ask me how I know.

It depends - read this, it addresses the sealing specification of tapered pipe threads, start about post #30.

Sorry, I am a lay-zee typist.

Never seen a pipe thread not leak without sealant, good luck with that.

No, it can not be liquid tight without sealant. The tapered threads still have helical leak paths at the roots and crests of the threads, because you can not cut a thread precisely enough that will contact in those corners at the same time that the thread faces meet.

Exceptions: plastic fittings (nylon, etc) can undergo enough plastic deformation to extrude into the thread roots to seal. A brass fitting is ductile enough that if it is way over-tightened, it may deform enough to achieve a seal. In doing so, you run the risk of splitting open the casing of the mating part.

The challenge here is that the floscan sensor has a very small internal passage with a little turbine in it. Any debris at all can obstruct the fuel flow, and/or interfere with the free rotation of that tiny turbine.

There has been at least one documented case of a tragic outcome from fuel obstruction in a fuel flow sensor, ( I believe from teflon tape, but I could be wrong)

What I did, and everyone can take this or leave it, it is simply what I did, was to use a very spare amount of teflon paste on the upper part of the threads, leaving a couple of threads completely clean. The hope was that whatever squeeze out of paste would mostly migrate outward, not inward, and even if a little bit did, it would move into those few bare threads and no farther.
After 500 hrs, no issues. I have not/can not look inside to see the actual state of intrusion of sealant into the sensor, but it works fine.

If the manufacturers of the sensor really want a liquid-tight connection with no sealant, they would machine a female "straight thread plus o-ring" boss, often called a JIC straight thread or SAE straight thread fitting. There is an AN spec for this and fittings are readily available. The fittings would have AN 37-degree flare male attachment on the other end, to receive a flare and B-nut. The cost difference between cutting a female pipe thread and a female AN o-ring boss thread is pretty small. It's just good engineering practice. Look at many many high pressure hydraulic devices to see how good connections are made that avoid use of sealant that can foul the internals.

Mine are fine 750 hours. I?ll research the threads attached and see what others are doing. Steel fittings into the Floscan, no issues.

Yes, you are wrong. Pipe threads are designed to self-seal. In the early days it was called pipe dope, not sealant because it is not designed as a sealant, but lubricant. In order to obtain a leak free NPT connection, it must be tight and requres a lot of force. There is a reason pipe wrenches are so long. The lubricant helps to achieve that at a much lower torque applied by the user and therefore increasing the likelihood of getting it tight enough. Most leakage is due to not understanding this and under-tightening joints.

Sealants are somewhat common in the plumbing trade because most/many threads are cut in the field with varying qualities of taps and dies. Sealants help when the threads have cracks and chips in them. That is not necessary with quality threads from a quality component.

I would assume the thread are NPTF, which is a unique cut with a very thin tip that easily deforms. I do believe that an NPT joint can seal liquid tight and have seen many examples of it. It does require proper intallation technique though. I also believe that slight deformation is typical, even in steel, and that is part of the design of the NPT thread. Have you noticed how it gets progressively tighter as you go? Then you back it out and run it in again, but it is much looser (i.e. easier to turn) until it hits the point where it was last resting? This is because the threads formed to one another via deformation. It is also why you need to go past the last tighten point if you disassemble the joint.

Also, NPT thread joints do seal at the roots. This is what the taper does. And what makes it so elegant is that all of that force eliminates the need for high precision of the thread. It forces the outside of the male thread into the root of the female thread and vice versa. The taper is forcing too big of a part into too small of a hole. All of that force you put on the joint is smashing that tip into the root and it deforms the tip until it fully complies. It is not elegant like a flare or ORB. It uses brute force to achieve it's goal. It is, however, time tested and proven when proper installation techniques are used.

All of that said, I always used permatex #2 on my NPT joints. I am far from perfect and this gives me some margin of protection from error as well as lubricates my threads. I also used at light coat on my red cube, but did also avoid the first 2 threads to be sure it didn't get into the sensor. #2 doesn't harden, so little bits can't break off and clog things.

Larry

While it is true that there are pipe threads designed to seal without using any thread sealant, regular NPT threaded fittings that we use on light aircraft are not one of them.

There is a lot of info available on line if someone is interested in learning more.
One resource I found doing a quick sear can be accessed HERE

I'm sorry Larry, but this is simply false. The taper forces the faces of the threads into intimate contact, and at that point, when the thread faces seat against each other, the roots and tips of the threads do not contact. All you have to do is look up the dimensional specs on the threads and draw them out to see that the root radius and tip radius are different so that when the thread faces meet, the tip of the male thread does not contact the root of the female thread.

And, as I said, if the manufacturer of the fuel flow sensor would use a proper straight thread with o-ring connection, we wouldn't even be having this discussion. See, for example, MS21900 for the appropriate male fitting to fit into a female straight-thread boss.