Thread Sealants and What They Are Actually Doing

Why this matters

A shop that had run tape on everything for years switched its vans to a paste, and inside two months it was replacing split female fittings on jobs where nothing else had changed. The techs had not changed how hard they pulled. The fittings had not changed suppliers. The sealant was doing something nobody had accounted for, and it was not the sealing.

That is worth understanding because it inverts the usual mental model. Most techs treat a thread sealant as the thing that stops the leak and the wrench as the thing that makes the joint tight. Both halves are only partly right, and the half nobody watches - what the sealant does to friction - is what breaks fittings.

What the thread by itself does and does not do

A tapered pipe thread is cut with clearance between the crest of one thread and the root of the other. That clearance is deliberate, it is how the threads assemble at all, and it forms a continuous helical channel from the inside of the pipe to the outside. The flanks of the threads bear on each other and generate the interference that makes the joint mechanically tight, but the spiral channel is still there.

So there are two separate things going on in a threaded joint:

• Mechanical tightness comes from flank interference as the taper drives the male into the female. That is what holds pressure and carries load.
• Sealing comes from something filling the helical clearance channel. On a machined, undamaged pair of threads that something can be metal-to-metal deformation alone, which is why some high-pressure joints are made up dry. On ordinary field threads it is the sealant.

A sealant that is doing its job is a gap filler in a channel roughly the width of a thread clearance. It is not holding pressure. It is not adding strength. It cannot bridge a cross-threaded start, a torn thread, a cracked casting, or a joint that has bottomed out with no interference left. If any of those is the actual condition, more sealant is a delay, not a repair.

The second job: friction

Here is the part that costs fittings.

A tapered thread does not make up to a position, it makes up to an interference. Every turn past hand-tight drives the male further into the female and increases the radial squeeze between them, which puts hoop tension in the female fitting. A sibling article works through that geometry in detail; what matters here is the direction: more turns, more hoop stress, and the female is where that stress lands.

Now, what does a wrench measure? Torque. And torque is split between the useful interference and the friction resisting motion at the thread flanks. Change the friction and you change how far the joint travels for the same torque, in the direction people find counterintuitive: a slicker interface means more turns for the same wrench effort, not fewer.

That is the whole mechanism behind the split fittings. Nobody pulled harder. The joints simply went further at the same pull.

The bench check that found it

Instead of arguing about it, the shop put the two sealants on the bench with a supply of the same fittings, marked hand-tight position with a paint line across the joint, made the joints up the way each tech normally would, and counted the turns past hand-tight.

Illustrative values from one shop's own fittings, not a spec. The ratio is the finding: 4.0 against 2.5 is 1.6 times the make-up, and 1.5 extra turns on a 14 threads-per-inch thread is 1.5 divided by 14, about 0.107 in of extra axial travel. On a taper that translates directly into extra radial interference in a fitting whose wall thickness did not change.

Three things follow from that table, and they explain field behavior the shop had been treating as unrelated:

The splits clustered on the thinner-walled and more brittle female parts. Same extra travel, same extra hoop stress, and the parts with the least margin went first. The tech was not doing anything different on those fittings; the fittings simply had less to give.

The failures were not immediate. A fitting stressed close to its limit at make-up can survive assembly and split later on a pressure cycle or a cold morning, which is why the pattern looked like a bad batch of fittings rather than a change in practice.

The joints that did not split leaked less. That is genuinely true and it is why the change had looked like an improvement at first. Extra make-up does improve sealing, right up until it does not.

What actually controls make-up

The control variable on a tapered thread is turns past hand-tight, not torque and not feel, and the acceptable range for a given size and material comes from the fitting manufacturer or the adopted plumbing, mechanical, or fuel gas code, not from a rule of thumb carried across materials.

Two practical consequences:

• Hand-tight is not a fixed reference across sealants. A paste that lubricates lets you thread further by hand before you feel resistance, which quietly moves the starting point of your count. Mark the joint at hand-tight with the sealant already on it, every time, not from a memory of where hand-tight used to be.
• Feel does not survive a change in materials. The same wrench effort on brass, on a plastic female thread, and on a malleable iron fitting produces three different outcomes, and the plastic one is where feel fails worst because it goes further before it protests and it splits without much warning.

The properties that actually differentiate the families

Different sealants differ in the leak path they fill, but the decisions that go wrong in the field are almost always about the other properties.

Cure behavior. Some sealants cure into a solid in the absence of air between metal parts, which also locks the joint and raises the breakaway torque at the next service. Others stay soft indefinitely, which keeps a joint serviceable and makes it more dependent on the mechanical interference staying put. A joint you know you will open every season and a joint that is buried in a wall are different service conditions with different right answers.

What it releases downstream. Tape that is applied over the first thread, or torn during make-up, sheds into the system. That debris finds strainers, cartridges, small orifices, and soft valve seats. This is why the practice of starting tape one or two threads back from the end exists, and it is not cosmetic.

Compatibility, which is a routing question, not a judgment call. Oxygen and other strong oxidizer service excludes hydrocarbon-based products outright, because that combination is an ignition source. Some plastics are attacked by solvent carriers or fail by environmental stress cracking with certain compounds. Fuel gas piping requires a sealant listed for that gas. Potable water requires certification for contact with drinking water, which reaches you through the plumbing code your authority having jurisdiction adopted rather than on its own. Every one of these belongs to the sealant manufacturer's data and the adopted code, and none of them can be resolved by looking at the joint.

Handling and heat

Do not apply heat to a joint that has thread tape or a cured sealant in it. PTFE decomposes at elevated temperature, roughly above 500 F, and the decomposition products are an inhalation hazard that produces a flu-like illness; solvent-carrying pastes give off vapor when heated. If a joint has to be brazed or soldered nearby, break the joint out and clean the threads first, work with ventilation that pulls fume away from your breathing zone, and follow your hot work permit and fire watch requirements under 29 CFR 1910.252 in general industry or 29 CFR 1926.352 in construction. Where your facility has adopted NFPA 51B, use the edition it adopted; it reaches you through that adoption or your contract, not on its own.

Wash hands before eating or smoking after handling tape. PTFE residue transferred to tobacco is a known route to the same fume exposure.

Read the safety data sheet for the specific product and use it, which is your right and your employer's duty under 29 CFR 1910.1200. Anaerobic products in particular have skin and eye provisions people ignore because the tube is small.

How to verify you got this right

• Count turns past hand-tight on a marked joint at least once for each sealant and each fitting material you routinely use. It takes one joint and it tells you what your own hands do.
• Inspect the female fitting at the hub after make-up, on brass, cast, and plastic parts especially. A fine circumferential line or a change in surface appearance at the deepest engagement is the beginning of a split, and it is visible before pressure finds it.
• Check what came off the joint at the next service. Tape shreds in a strainer are a record of how the last joint was made up.
• When fittings start failing after nothing changed, ask what consumable changed. Sealant, lubricant, and tape brand and grade are the invisible variables in a shop, and none of them appear on a work order.

References

• 29 CFR 1910.1200, hazard communication, for access to and use of the safety data sheet for the specific product
• 29 CFR 1910.252 (general industry) and 29 CFR 1926.352 (construction) for hot work fire prevention where a joint is heated, with NFPA 51B in the edition your facility has adopted
• Fitting manufacturer documentation for permitted turns past hand-tight by size and material; sealant manufacturer documentation for temperature, pressure, and material compatibility
• Plumbing and fuel gas codes as adopted by your authority having jurisdiction, for potable water certification and fuel gas listing of sealants
• See related: Why a Tapered Thread Seals and a Straight One Does Not; The Thread Sealants and Where Each One Belongs; Cross-Threading Prevention