Purge Dams for Every Pipe, Tube and Weld Joint

A purge dam is a temporary barrier inserted inside a pipe, tube, or vessel on either side of a weld joint to seal off a small purge chamber around the root. Inert gas, almost always argon or nitrogen, is fed into that sealed section to displace the air, so the back (root) side of the weld is shielded from oxygen while the weld is made. Without a dam, the operator would have to purge the entire length of pipe, which wastes gas and takes far longer to reach a safe oxygen level.

On stainless steel, duplex, titanium, and nickel alloys, the molten weld pool and the hot metal immediately around it will react with any oxygen present, on the top (face) side or the underside (root). The torch and shielding gas protect the face, but the root is enclosed inside the pipe and gets no shielding unless the atmosphere inside the pipe is purged. An unprotected root discolors and oxidizes, which is the visible sign of a much more important problem: the surface has lost the chromium oxide layer that gives the alloy its corrosion resistance.

Sugaring is the industry term for the granular, crystalline oxide scale that forms on an unpurged or under-purged stainless root pass; it gets its name because the texture resembles coarse sugar. It is a visible symptom of oxidation, not just a cosmetic issue: the scale itself and the depleted chromium layer underneath it are both less corrosion-resistant than the surrounding parent metal, which is exactly the property the alloy was specified for in the first place. In sanitary and high-purity service, sugared roots are also harder to clean and can harbor contamination.

It depends on the alloy and the service. As a general rule of thumb widely used in the industry, oxygen below roughly 50 ppm produces a bright, straw-free root on standard austenitic stainless steel, while oxygen in the low hundreds of ppm starts to show light straw or gold discoloration, and levels in the thousands of ppm produce visible sugaring. More demanding materials and services, titanium, some nickel alloys, and high-purity semiconductor or pharma lines, commonly call for tighter targets, sometimes under 20-30 ppm or lower. Always confirm the actual number against your welding procedure specification (WPS), the governing code, or the customer spec rather than relying on a rule of thumb alone.

Many do, for certain material classes and services. Process piping codes such as ASME B31.3 call for inert gas backing on root passes for materials susceptible to oxidation, and sanitary/hygienic tube welding standards such as AWS D18.1 set specific purge oxygen requirements for the finished weld. Nuclear, semiconductor, and aerospace specifications frequently add their own, tighter oxygen ceilings. Check the WPS and the governing code or customer specification for the job; requirements vary by material, wall thickness, and service.

The main families are water-soluble paper or film dams (dissolve away after use, no retrieval needed), inflatable bladder dams (elastomer bladders inflated with air or water to seal the pipe ID, used single-ended or as a matched pair to bracket the joint), foam or plug dams (a compressible foam disc or plug pushed into position), rigid or expandable mechanical disc dams (a solid or expanding metal/composite disc that seats against the ID), and self-adhesive pre-formed dams (a flexible pad with an adhesive edge pressed against the pipe wall). Each trades off differently on removability, reusability, temperature tolerance, and how much access you need to install and retrieve it.

Start with the pipe or tube inside diameter, every dam type has a rated ID range, and confirm the dam actually fits before it ships to the shop. From there, weigh how much you need the chamber volume reduced (a tighter-sealing dam purges faster on a large-diameter run), how close the dam sits to the weld and what temperature it will see, how it needs to come out afterward (dissolve in water, deflate and pull, or physically retrieve), whether it is reusable across many joints or single-use, and whether you have access from one end of the pipe or both. A water line or double-ended pipe run needs a different answer than a single dead-end nozzle.

On a long pipe run, placing one dam on each side of the weld joint, rather than purging from one dam all the way to an open pipe end, brackets off a small, fixed-volume chamber right around the root. That small chamber purges to a safe oxygen level in a fraction of the time and gas volume a full-length purge would need, which is the main reason purge dams exist in the first place. Both dams need a way to vent displaced air out as the inert gas comes in, otherwise the sealed chamber pressurizes instead of purging.

There is no single number; it depends on the purge chamber volume (which is why dams matter), the gas flow rate, the inlet and vent arrangement, and the target oxygen level. In practice, the operator sets a flow rate suited to the chamber size and dam vent capacity, watches an oxygen analyzer (a weld purge monitor) reading gas drawn from inside the chamber, and holds the purge until the reading is at or below the target before striking an arc, then keeps purging at a reduced flow through the pass. Judging purge time by the clock alone, without a monitor reading, is a common and risky shortcut.

Every dam setup needs a controlled vent or outlet path for the gas and displaced air to leave the sealed chamber; purging into a fully sealed section with no vent will bulge or blow a dam, and on a thin-wall tube it can distort or even bulge the root pass itself. Inflatable bladder dams usually have a separate purge-gas port distinct from the inflation port precisely so the chamber has a vent path; foam, mechanical, and adhesive dams need a deliberate small gap, vent tube, or bleed hole built into the setup. Confirm the vent path before starting the purge, not after noticing a pressure problem.

It is designed to be left in place rather than retrieved: after the weld is complete, water (or in some cases a solvent, per the manufacturer's instructions) is flushed through the line and the paper or film dissolves and washes out. This makes water-soluble dams well suited to piping that will be flushed or hydrotested anyway, and to configurations where a mechanical dam could not physically be pulled back out, such as a joint deep inside a long, small-ID run.

It depends on the type. Inflatable bladder dams and rigid or expandable mechanical dams are generally reusable across many joints, provided the elastomer or seal surface is inspected and not damaged by heat or handling. Water-soluble dams and most self-adhesive pre-formed dams are single-use by design, they either dissolve away or lose their seal once removed. Foam/plug dams vary by construction; check the specific product's rating before planning to reuse one.

Yes. Orbital GTAW on small-diameter tube, common in semiconductor ultra-high-purity gas lines, pharmaceutical and sanitary process tubing, and aerospace tube runs, depends on a tightly controlled root-side atmosphere to meet demanding oxygen and surface-finish specifications, and dams are the standard way to keep the purged volume small enough to reach that target quickly on a production line. Pre-formed and inflatable dams sized for small tube ODs are common in this setting, often used in pairs to bracket a single fixed joint on an orbital head.

They do two different jobs and most root-critical welds use both. A purge dam is a physical barrier that seals off the chamber so a small purge volume can be flushed and held; it controls gas, not the reading of it. A weld purge monitor is an instrument, typically drawing a gas sample from the chamber to an oxygen sensor, that tells the welder the actual oxygen level inside the sealed volume so they know when it is safe to strike an arc and whether it stays in range during the pass. A dam without a monitor is a guess; a monitor without a dam has nothing small enough to measure quickly.

Two mistakes come up most often: sizing the dam to the wrong pipe ID so it does not seal fully around the circumference, letting outside air leak back in during the weld, and starting to weld on a time estimate instead of an actual oxygen reading from a purge monitor. Both produce a root that looks acceptable on the outside but reads out of spec once the reading, or the surface, is checked. Confirming fit before the job and purging to a verified reading rather than a guess avoids the great majority of dam-related rework.