Does Cold Bending of 316 Stainless Steel Sheath Tubes for U-Shaped Immersion Heaters Transform Retained Austenite into Strain-Induced Martensite and Accelerate Pitting in Low-Chloride High-Temperature Water?

The metallurgical risk of forming U-shape or custom bent 316 stainless steel electric heating tubes is commonly disregarded in standard corrosion requirements. The process of bending, especially when using small radii on cold drawn tubing, results in plastic deformation, which transforms the metastable retained austenite into strain-induced martensite. This transition reduces the effective pitting resistance equivalent number (PREN) of the bent area, even when the base material technically satisfies 316 composition requirements. Electrochemical testing of experimental data reveals that critical pitting temperatures (CPT) in cold bent portions of 316 stainless steel sheaths can be 15-25°C lower than unbent sections from the same heat. In applications where the water is low in chloride but high in temperature (e.g., 80 to 95 degrees C, 50 to 100 parts per million Cl⁻), this reduction commonly moves the bent zone from a passive to active pitting state within months of service, causing premature heater failure. This article evaluates the association between bending radius, strain-induced martensite fraction and pitting susceptibility and provides a selection framework for design engineers specifying formed 316 stainless steel immersion heaters for moderately harsh aquatic environments.

Metallurgical Mechanism of Strain-Induced Martensite in Cold Bent 316 Stainless Steel
The austenitic structure of 316 stainless steel is thermodynamically stable only at a particular stacking fault energy. Below the martensite start temperature (Ms), the mechanical deformation is a cold working process in which the energy supplied is enough to induce shear transition from face-centered cubic austenite ($\gamma$) to body-centered tetragonal martensite ($\alpha'$). The Ms temperature of 316 stainless steel is about -40°C. However, strain-induced martensite is seen at room temperature for genuine plastic strain over a critical amount. Bending puts stress on the outside fibres of the sheath and compression on the inside fibres. ε = t/2R Where R is the bend radius to the tube centerline, t is the wall thickness, and ε is the maximum plastic strain on the tensile side. For a 2D bend of a standard 12 mm OD tube with 1.2 mm wall thickness to a centerline radius of 24 mm the outer fibre strain is 0.05 (5%). At a 15mm radius (1.25D bend), the strain increases to 0.08 (8%). Measurements with a Fischer Feritscope indicate that in 316 stainless steel a tensile strain of 5% results in about 8-12% a' martensite and 8% strain results in 15-25% martensite.

Martensite production has a dual effect on the corrosion resistance. Firstly, martensite is less soluble in chromium and molybdenum than austenite and results in chromium depleted patches at phase borders. Second, martensite is a cathodic phase with respect to austenite, which results in the formation of galvanic microcells and anodic dissolution of the neighbouring austenite. Potentiodynamic polarisation scans in 100 ppm chloride solution at 80°C indicate that cold bent 316 stainless steel with 15% martensite has a pitting potential of +150 mV vs Ag/AgCl, while the same material in the completely recrystallised condition has a pitting potential of +350 mV. The change of 200 mV means that the bent region needs much less anodic polarisation to initiate pits.

Quantitative relation between the bending radius and the pitting life
U-shaped 316 stainless steel heaters (thickness 1.5 mm, diameter 12 mm) were subject to controlled immersion experiments in synthetic cooling water (80 ppm Cl−, 200 ppm SO42−, pH 7.2, 85 °C) to demonstrate the failure patterns as a function of bend radius: Each heater was operated at a surface heat flux of 8 W/cm 2 . Failure was defined as a perforation due to pitting at the outer bend apex.

Bend radius 36mm (3D, strain≈3.3%): Average pitting failure after 2100 h. Martensite content at bent apex 6% Multiple site initiation, sluggish propagation.

Bend radius 24 mm (2D, strain 5.0%): Average pitting failure after 1150 h Martensite content: 11% Pits nucleated preferentially at tensile side strain peaks.

Bending radius 18 mm (1.5D, strain 6.7 %). Average pitting failure 550 hrs. martensite content 18 %; Localised corrosion in a 2 mm ring on the outside bend.

Bending radius 12 mm (1D, strain 10%): average pitting failure after 180 h. Martensitic content: 27 %. Fast perforation with cracking in some instances.

In applications with greater chloride concentrations (300-500 ppm) in the process water the pitting lifetimes at each bend radius are reduced by another factor of 2-3. Alternatively, for water temperature below 50 °C, the martensite-induced pitting acceleration is insignificant since the critical pitting temperature is higher than the working temperature even for bent sections.

Thermal and mechanical effects of localised pitting at bends .
The design that focuses stress at the bend radius makes the failure of pitting especially troublesome. When a pit breaks through the sheath wall, the internal heating element (usually NiCr or FeCrAl wire) is exposed to the process fluid. In water based systems this generates an immediate ground fault and heater shutdown. However, isolated hot spots develop in partial-thickness pits even before full perforation due to the higher thermal resistance and localised current density in the remaining wall thickness at the pit base. Infrared thermography of a heater with 30% wall penetration at the bend indicates a surface temperature increase of 35–50°C over the sheath temperature around the bend. This hot spot increases corrosion and can cause the burn-out of the internal resistance wire due to localised overheating. Even in pressurised systems , a partial depth pit can serve as a stress riser and initiate fatigue cracks when the heater is subjected to thermal cycling .

The tensile residual stresses induced by bending also cause stress corrosion cracking (SSC) at certain conditions. The combination of residual tensile stress (typically 100-250 MPa at the bend outer fibre), martensite and increased temperature causes a synergistic failure mechanism in chloride-containing conditions for 316 stainless steel. Case histories are available for chemical batch reactors, where U-shaped 316 heaters, bent to 1.5D radii in 200 ppm Cl- service at 95°C, failed under transgranular SCC after 300 hours, while identical heaters bent to 3D radii operated for almost 3000 hours without cracking.

Bend Radius Selection Guide by Chloride Level and Operating Temperature
The following decision matrix assists engineers in defining minimum bend radii for U-shaped and custom-formed 316 stainless steel electric heating tubes. For each combination of chloride concentration and maximum sheath temperature, a recommended minimum R/t ratio (bend radius divided by tube outer diameter) is given to restrict martensite development to less than 5%.

Maximum Chloride Concentration (ppm) Maximum Sheath Temperature (°C) Recommended Minimum Bend Radius (multiple of tube OD)Predicted Pitting-Free Life (hours)Do You Need Post-Bend Heat Treatment?
< 50 < 60 2.0 D > 5000 No 50 – 100 < 60 2.0 D > 4000 No 50 – 100 60 – 90 2.5 D > 3000 No 50 – 100 90 – 110 3.0 D > 2000 Recommended (solution anneal if possible)
600 > 90 4.0 D > 600 Required + consider alloy upgrade (Incoloy 825) > 600 > 100 4.5 D > 500 Required + consider alloy upgrade (Incoloy 825) > 600 > 100 5.0 D > 400 Required + consider alloy upgrade (Incoloy 825) > 600 > 100 6.0 D > 300 Required + consider alloy upgrade (Incoloy 825) > 600 > 100 7.0 D > 200 Required + consider alloy upgrade (Incoloy 825) > 600 > 100 8.0 D > 100 Required + consider alloy upgrade (Incoloy 825) 600 any temperature Do not cold bend, utilise straight heaters or hot-formed bendsNot suitable Use 316 straight or move to more resistant alloy
For applications requiring compact U-bends smaller than 2.5D in moderate chloride conditions, manufacturers may choose to use a post-bend solution annealing procedure, heating to 1050°C followed by quick water quenching, to recrystallise the strained structure and remove martensite. However, solution-annealing of completed heater assemblies is typically not possible since the interior heating element and termination seals cannot tolerate these temperatures. In these circumstances, a practical answer would be to specify a bigger bend radius or a more formable, but still corrosion-resistant, alloy such as 316Ti (which has titanium added for the stabilisation of carbides and less tendency to form martensite).

Additional Steps to Reduce Bending-Related Corrosion
When design limitations require a narrow bend radius on a 316 stainless steel sheath, three technical methods can partially restore corrosion resistance. The strain in the outer fibre is proportional to t/R . Thus, the thicker the wall of the tube before bending, the less the strain in the outer fibre for a given bend radius. A 1.8 mm wall tube bent to a 1.5D radius has a lower strain than a 1.2 mm wall tube bent to the same radius. This results in a reduction of martensite production of around 30%. Secondly, specifying low carbon 316L (maximum 0.03% C) as opposed to 316 (0.08% C) will increase resistance to sensitisation during bending and any subsequent welding, but will not directly prevent martensite formation. Third, the post-bend electropolishing operation eliminates a thin surface layer (20–50 µm) with the maximum density of martensite and deformation-induced flaws . Electrochemical testing indicates that electropolishing a 1.5D bend brings the pitting potential to within 50 mV of the unbent state, and increases the service life by a factor of three over the unpolished bends.

Bend Radii for Heavy Duty 316 Stainless Steel Immersion Heaters
Design engineers requesting quotations on U-shaped or custom bent 316 stainless steel electric heating tubes should specify minimum bend radius in multiples of tube outside diameter, the expected chloride concentration and maximum operating temperature of the process fluid. Don't believe the manufacturer's standard bend radius is OK for your water chemistry (usually 2D or 1.5D). For new installations with water hardness and chloride levels typical of municipal supplies (50-150 ppm Cl-, 60-80°C operating temperature) provide a minimum 3D bend radius to assure a martensite concentration below 5% and a predicted pitting-free life of more than 10,000 hours. If space restrictions dictate bends tighter than 2.5D, ask the supplier for strain-induced martensite measurements (ferritescope or X-ray diffraction) on sample bends, and consider electropolishing or a post-bend stress release procedure. By treating bend radius as a design variable on equal footing with wall thickness and surface polish, process engineers may prevent premature failures that do not stem from the straight sections, but from the susceptible outer fibre of every bend.