Seamless tube making is the hot-rolling process that converts a solid round steel billet into a hollow tube without any longitudinal weld seam. The core component is the Mannesmann rotary piercing mill — two barrel-shaped skew rolls that spin the billet against a hardened plug, opening a central cavity along the bar's axis. The process exists because welded pipe cannot be trusted at the pressures and temperatures seen in oil-country tubular goods, boiler superheaters, and aerospace hydraulic lines. A typical mill turns a 200 mm billet into 50 m of seamless casing in under 8 minutes.
Seamless Tube Making Interactive Calculator
Vary billet size, shell geometry, spread factor, and billet length to see the Mannesmann piercing elongation and shell length.
Equation Used
The elongation ratio compares the pierced hollow shell length with the starting billet length. It is estimated from conservation of cross-sectional metal area: the solid billet area is divided by the shell annulus area, then corrected for spread factor S. Typical Mannesmann piercing targets are about 2.5 to 3.5.
- Hot steel volume is approximately conserved through the piercing pass.
- Billet and pierced shell are treated as circular sections.
- Spread factor S represents fractional OD growth caused by skew rolling.
- Formula estimates piercing elongation only, not downstream mandrel or stretch-reducing mill changes.
Inside the Seamless Tube Making
The starting stock is a hot round billet — usually 150 to 400 mm diameter, heated to roughly 1200 °C in a rotary hearth furnace. You feed that billet between two barrel rolls set on skew axes, typically with a 3° to 12° feed angle and a 0° to 10° cross angle. The rolls spin in the same direction, but because they are tilted, the billet rotates AND advances at the same time. As the steel rotates under heavy radial compression, the centreline goes into tensile stress — the metal literally pulls itself apart along the axis. That's the Mannesmann effect, discovered by the Mannesmann brothers in 1886, and it is what lets a hardened plug positioned on a mandrel rod open a clean central bore without drilling.
Why design it this way? Because drilling a 12 m long axial hole in solid steel is impossible at production rates, and welding a longitudinal seam leaves a heat-affected zone that fails first under pressure cycling. Hot piercing produces a hollow shell with continuous, undisturbed grain flow wrapping the bore — exactly what an OCTG seamless pipe needs to survive 15,000 psi downhole. After piercing, the hollow shell goes to a plug mill, mandrel mill, or Assel mill for wall reduction, then a stretch-reducing mill sets the final outside diameter.
Tolerances matter brutally here. If the plug nose sits more than ±2 mm off the roll-gorge centreline, you get eccentric wall thickness — one side of the finished tube reads 8.0 mm and the other reads 6.4 mm, and the pipe fails ultrasonic inspection at the mill exit. If the billet temperature drops below about 1150 °C before piercing, the centreline tensile stress can't open cleanly and you get internal laps and slivers that look like fish scales on the bore. If the feed angle is set too shallow, throughput collapses; too steep and the rolls bite chunks out of the billet skin. A real plant runs the rolls within ±0.5° of the design feed angle, every shift.
Key Components
- Barrel-shaped skew rolls: Two driven rolls, typically 1000 to 1400 mm diameter, mounted on axes tilted at the feed and cross angles. They rotate the billet while advancing it through the gorge. Roll material is forged 60CrMoV steel and a fresh pair holds shape for around 800 to 1500 piercing passes before re-machining.
- Piercing plug: A bullet-nose tool, usually molybdenum-alloy or chrome-molybdenum, sitting on the end of a mandrel bar at the roll gorge. The plug is water-cooled internally and lasts 30 to 200 piercings depending on grade — a P110 OCTG steel eats plugs faster than a low-carbon line-pipe grade.
- Mandrel bar: A long, water-cooled rod that holds the plug in position against the axial thrust of the advancing billet. Bar straightness must hold within 0.5 mm/m or the bore wanders. On a typical retained-mandrel mill the bar is 25 to 35 m long.
- Centring guides (Diescher discs or shoes): Idle discs or fixed shoes between the rolls that constrain the billet on the unrolled axis. Without these the billet spirals out of the gorge. Disc edge clearance to the billet must hold within ±1 mm.
- Rotary hearth furnace: Reheats billets to 1200 to 1280 °C with a soak time of 2 to 4 hours. Temperature uniformity end-to-end must stay within ±15 °C — any colder spot pierces with internal defects.
- Plug or mandrel mill (downstream): Reduces wall thickness of the hollow shell from roughly 30 to 50 mm down to the finish wall of 5 to 25 mm. A retained-mandrel mill (MPM) holds wall tolerance to ±8% on premium product.
- Stretch-reducing mill: A train of 12 to 28 small 3-roll stands that pulls the tube down to final OD without an internal tool. Each stand reduces OD by 1 to 5%, and the cumulative draw can stretch a 12 m hollow into a 100 m finished tube.
Industries That Rely on the Seamless Tube Making
Seamless tube exists wherever a welded seam would fail — high pressure, high temperature, fatigue cycling, sour service, or aerospace certification. The big tonnage goes into OCTG (oil-country tubular goods), boiler and superheater tubes, hydraulic cylinder barrels, and mechanical tubing for driveshafts. Smaller volumes feed aerospace hydraulic lines, nuclear steam generator tubing, and high-pressure gas cylinders. You see seamless specified anywhere a welded ERW pipe would crack at the bond line under cyclic load.
- Oil and gas: Tenaris TenarisHydril 5-1/2 in P110 casing produced on the Dalmine PQF mandrel mill for deepwater wells in the Gulf of Mexico
- Power generation: Vallourec T91 superheater tubing for ultra-supercritical boilers at RWE's Niederaussem unit, 50 mm OD × 7.5 mm wall
- Aerospace hydraulics: Sandvik 21-6-9 stainless seamless tubing for Airbus A350 hydraulic systems, certified to AMS 5561 at 5,000 psi working pressure
- Mechanical engineering: Mannesmann DMV cold-drawn seamless tube used as hydraulic cylinder barrels in Enerpac RCH-series hollow-plunger cylinders
- Nuclear: Sumitomo SMI Inconel 690TT steam generator tubing for AP1000 reactors, 19.05 mm OD × 1.09 mm wall on a Pilger-rolled finish
- Automotive driveline: Benteler seamless DOM tube used as propshaft stock for Ford F-150 rear driveshafts, 76 mm OD × 2.5 mm wall
- Gas storage: DOT-3AA seamless steel cylinders rolled from billet on a Mannesmann Erhardt push bench for industrial oxygen service at 2,265 psi
The Formula Behind the Seamless Tube Making
The single most useful formula in seamless tube making is the elongation ratio of the piercing pass — how much longer the hollow shell becomes compared to the starting billet. This number tells you whether the pierce is gentle enough to hold centreline integrity or so aggressive it tears the bore. Run too low (below about 1.8) and you waste billet, throughput suffers, and the wall comes out thick and sluggish to roll downstream. Run too high (above about 4.5) and centreline porosity, internal laps, and plug overload start showing up on UT inspection. The sweet spot for most carbon and low-alloy grades sits between 2.5 and 3.5.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| λ | Elongation ratio of the piercing pass (dimensionless) | — | — |
| Lshell | Length of the hollow shell after piercing | m | ft |
| Lbillet | Length of the starting billet | m | ft |
| Dbillet | Outside diameter of the round billet | mm | in |
| Dshell | Outside diameter of the pierced hollow shell | mm | in |
| dbore | Inside (bore) diameter set by the plug | mm | in |
| S | Spread factor — fractional OD growth caused by skew rolling, typically 0.04 to 0.08 | — | — |
Worked Example: Seamless Tube Making in a Mannesmann piercing mill rolling P110 casing billets
A seamless tube plant in Changzhou is piercing 220 mm round P110 billets at 1230 °C on an SMS Meer cone-type piercing mill. Target hollow shell is 235 mm OD with a 165 mm bore set by a chrome-moly plug. Spread factor measured on this pass schedule is 0.06. Starting billet length is 2.40 m. The shift supervisor wants to know the elongation ratio at the nominal pass and what happens if the plug retraction control drifts the bore diameter at the low and high ends of the typical operating window.
Given
- Dbillet = 220 mm
- Dshell = 235 mm
- dbore,nom = 165 mm
- Lbillet = 2.40 m
- S = 0.06 —
Solution
Step 1 — compute the cross-sectional area ratio at the nominal 165 mm bore. The billet is solid, the shell is an annulus:
Step 2 — apply the spread correction and compute nominal elongation ratio:
So the 2.40 m billet exits the piercer as a 2.40 × 1.84 = 4.42 m hollow shell. That's gentle — wall is heavy, bore is clean, and the hollow goes to the mandrel mill with plenty of meat for the wall draw.
Step 3 — at the low end of the typical operating window, the plug retracts and the bore opens to 175 mm:
Shell length climbs to 5.02 m — still well inside the safe envelope for P110.
Step 4 — at the high end, the plug advances hard and the bore closes to 150 mm:
Now you're under-piercing. Shell comes out at 3.77 m with a heavy 42.5 mm wall, the mandrel mill struggles to draw it down in the available stand count, and you'll see roll motor amps spike on stands 4 and 5 of the MPM. You would be amazed how often plant supervisors blame the mandrel mill for wall variation that actually started at the piercer plug position.
Result
At the nominal 165 mm bore the elongation ratio is 1. 84 — the 2.40 m billet becomes a 4.42 m hollow shell. In the rolling pulpit that feels like a clean, quiet pierce with steady main-drive amps and no audible chatter. Drop the bore to 150 mm and λ falls to 1.57 with a heavy-walled 3.77 m shell that overloads the downstream MPM; open the bore to 175 mm and λ climbs to 2.09 with a 5.02 m shell that runs easily but eats plug life. If your measured shell length differs from the predicted value by more than ±5%, the most common causes are: (1) plug nose worn back by 3 to 8 mm so the effective bore is larger than the nominal setpoint, (2) billet temperature variation across the length giving differential spread — a cold tail end pierces stiffer and shorter, or (3) Diescher disc clearance opened past 2 mm letting the billet wander off the gorge centreline, which shows up as spiral wall variation on the UT trace.
When to Use a Seamless Tube Making and When Not To
Seamless tube is not always the right answer. Welded ERW pipe is cheaper, faster, and dimensionally tighter for the same money — the question is whether your service condition forgives a longitudinal weld. Here is how the three main tube-forming routes line up on the dimensions buyers actually compare.
| Property | Seamless (Mannesmann) | ERW welded | Spiral / SAW welded |
|---|---|---|---|
| Maximum working pressure (typical) | Up to 20,000 psi (OCTG, hydraulic) | Up to 3,000 psi (line pipe) | Up to 1,500 psi (large-diameter transmission) |
| OD range produced economically | 20 to 660 mm | 10 to 660 mm | 300 to 2,500 mm |
| Wall tolerance (best case) | ±8% on retained-mandrel mill | ±5% on continuous ERW | ±0.5 mm on plate |
| Cost per tonne (carbon steel, 2024) | $1,400 to $2,200 | $800 to $1,300 | $900 to $1,400 |
| Fatigue life under pressure cycling | Excellent — no seam, continuous grain flow | Moderate — bond line is the failure initiator | Moderate — spiral seam adds stress concentration |
| Suitability for sour service (NACE MR0175) | Standard route for H2S service | Restricted — requires special HFI process | Generally unsuitable |
| Plant capital cost | $200 to $500 million for a full mill | $30 to $80 million | $20 to $60 million |
| Throughput per hour | 20 to 80 tonnes/hr | 60 to 150 tonnes/hr | 10 to 40 tonnes/hr |
Frequently Asked Questions About Seamless Tube Making
Centreline eccentricity in a piercer almost never comes from the plug itself. The usual cause is unequal billet rotation speed at the entry side versus the exit side of the gorge — which is set by the cross angle of the rolls, not the feed angle. If the cross angle on the top roll drifts 1° away from the bottom roll, the billet rotates faster on one side as it passes through and the plug cuts deeper on the slower-rotating side.
Check the cross-angle synchronisation on the roll housings before you blame the plug or the mandrel bar. A good diagnostic is to mark the billet end with chalk before entry and watch the helical track of the chalk on the shell — uneven helix pitch means uneven rotation.
Plug mill (two-stand reversing) is the old route — flexible, slow, good for short runs and exotic alloys where you'd rather change tooling than commit to a long mandrel campaign. Mandrel mill (continuous MPM or PQF, 5 to 8 stands) is the high-volume choice for OCTG and line pipe, holding wall tolerance around ±8% and pumping out 60+ tonnes/hour. Assel mill is the specialist — three skew rolls, used for thick-wall bearing-quality and mechanical tubing where you need a very clean ID surface and a length-to-diameter ratio under 12.
Rule of thumb: under 100 km/year of any single size, plug mill. Over 300 km/year, mandrel mill. Bearing races, gun barrels, hydraulic cylinder stock — Assel.
P110 has higher chromium and molybdenum content and roughly twice the hot yield strength of J55 at piercing temperature. The plug nose sees a higher contact stress against the harder steel, and the heat flow into the plug body climbs because the deformation work per unit volume is larger. A plug that survives 200 J55 piercings will typically last 30 to 60 P110 piercings.
The fix is not a harder plug material — it's better plug cooling. Increase the internal water flow, drop the inlet temperature below 25 °C, and consider switching to a Mo-W-alloy plug with a thermally sprayed scale-resistant coating. Also pre-heat your billets uniformly — a hot spot on the billet of even +30 °C will pull a divot out of the plug nose on a single pass.
Stretch-reducing mills (SRM) pull the tube down in OD without an internal tool, using inter-stand tension to also reduce wall — you can hit a dozen finish sizes from one mother-tube size, which is gold for product mix. The catch is that wall thickens at the tube ends (the off-gauge ends) because the first and last stands don't have the full tension sequence; you crop and scrap that material.
Sizing mills only adjust OD by 1 to 3% and don't change wall meaningfully. Pick a sizing mill if your product mix is narrow and you want minimum end-crop loss. Pick an SRM if you're running 20+ size combinations off the same hollow shell and you can absorb the 3 to 6% end-crop yield hit.
Internal slivers in a seamless tube almost always trace back to the piercer, not the mandrel mill. The two dominant root causes are: billet centreline porosity (the as-cast billet had shrinkage voids on the axis that opened up under the Mannesmann tensile stress), and entry-side billet skin defects that got rolled inward during piercing because the billet was too cold at the entry end.
Diagnostic check: pull a longitudinal section through one of the slivered tubes and look at the orientation. Slivers that spiral around the bore at the same helix angle as the piercer rotation are billet-origin defects. Slivers that run axially are mandrel-bar scoring — different problem entirely. Push back on your billet supplier with a centreline soundness spec (ASTM E381 macroetch grade) before you change anything on the mill.
The original cavity in the billet always requires hot work — you cannot open a centreline bore at room temperature in solid steel without either drilling (uneconomic) or extrusion (different process, limited to short lengths and softer alloys). What you can do is finish a hot-pierced mother tube by cold drawing or cold pilgering. Sandvik and Tubacex routinely cold-pilger stainless and nickel-alloy seamless tubing to OD tolerances of ±0.05 mm and Ra under 0.4 µm on the bore.
So the answer is: hot pierce is mandatory; cold finishing is optional and is what gets you to aerospace and instrumentation tolerances.
Two effects almost always explain a 3 to 5% length shortfall. First, end crops — both the front nose and the tail of the hollow shell are out-of-round and off-gauge after piercing, and the shear cuts off 200 to 600 mm at each end before the shell enters the mandrel mill. That alone explains 2 to 4% on a 4 m shell.
Second, your spread factor S in the formula may be set too high for your actual pass schedule. S depends on the cross angle, feed angle, and reduction ratio — measured spread on a real cone-piercer is often 0.04 rather than the 0.06 to 0.08 quoted in textbooks. Re-measure Dshell on three pierced shells with calipers at the gorge exit and back-calculate S before you trust the formula.
References & Further Reading
- Wikipedia contributors. Mannesmann process. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
