Bellows Stiffness Testing: Determination of Axial, Lateral, Angular and Ring Stiffness of Metallic B

Bellows stiffness testing quantifies the spring rate of metallic bellows and corrugated pipes — the force or moment required to produce a unit axial, lateral, or angular displacement — and is essential for verifying piping loads, movement capacity, and fatigue performance. This article explains the test principles, sample requirements, measurement procedures, and the applicable standards such as EJMA, ASME B31.3, EN 14917, GB/T 12777, ISO 9969, and ASTM D2412. It is intended for design engineers, quality engineers, and procurement specialists who need reliable stiffness data for expansion joints, bellows components, and corrugated piping products.

Technical Overview

Bellows stiffness, commonly referred to as spring rate, is defined as the force (or moment) required to produce a unit displacement of a bellows element. It is a primary design characteristic of metallic bellows expansion joints and corrugated piping products because it directly determines the loads transmitted to pipe anchors, guides, equipment nozzles, and connected structures. Insufficient or excessive stiffness can lead to overstress, fatigue cracking, vibration problems, or unacceptable nozzle loads.

Depending on the loading direction, the following stiffness quantities are distinguished:

• Axial stiffness: axial force per unit axial displacement, expressed in N/mm (lbf/in), measured in both compression and extension.
• Lateral (transverse) stiffness: lateral shear force per unit lateral offset, expressed in N/mm.
• Angular (bending) stiffness: bending moment per unit angular rotation, expressed in N·m/degree (or per radian).
• Torsional stiffness: torque per unit twist angle about the bellows axis, tested only when explicitly specified, since most bellows are not designed to absorb torsion.
• Ring stiffness (SN): the diametric load resistance of thermoplastic corrugated pipes, expressed in kN/m², which governs the deflection behavior of buried pipes under soil and traffic loads.

For metallic bellows, the governing references are the EJMA Standards of the Expansion Joint Manufacturers Association, ASME B31.3 (including the expansion joint appendix), EN 14917, and GB/T 12777. These documents provide theoretical stiffness calculation models as well as requirements for type testing and production testing. For thermoplastic corrugated pipes, ring stiffness is determined in accordance with ISO 9969, GB/T 9647, and ASTM D2412 using the parallel-plate loading method. Accredited third-party laboratories routinely perform these determinations for design verification, product certification, and failure analysis.

Testing Samples

Metallic Bellows and Expansion Joint Assemblies

Test articles include free-length bellows elements and complete expansion joint assemblies, both reinforced and unreinforced. Typical materials are austenitic stainless steels such as grades 304, 316L, 321, and 347, nickel alloys such as Alloy 625, Alloy 825, and Monel 400, and specialty alloys for high-temperature or corrosive service. Samples may be hydroformed, mechanically (roll) formed, or welded (edge-welded) bellows, in single-ply or multi-ply construction, covering nominal sizes from small instrument bellows (DN 8) to large-diameter power and petrochemical expansion joints (DN 3000 and above).

Thermoplastic Corrugated Pipes

Structured-wall products such as HDPE, PVC, and polypropylene double-wall or triple-wall corrugated pipes are tested for ring stiffness. These include buried drainage and sewer pipes, storm-water pipes, and corrugated cable protection conduits, commonly classified by ring stiffness classes such as SN2, SN4, and SN8.

Sample Preparation and Conditioning

Bellows samples are supplied at free length with the specified convolution count, with end fittings welded or flanged as required for clamping. Samples are identified by heat number and manufacturing lot, and geometric parameters — convolution pitch, height, wall thickness per ply, number of plies, and number of convolutions — are verified before testing. Thermoplastic specimens are cut to the standard length (typically 300 mm ± 10 mm) and conditioned at 23 °C ± 2 °C in accordance with ISO 291 before loading. For metallic bellows, testing is normally performed at ambient temperature, with the test temperature recorded.

Testing Items

The scope of a bellows stiffness determination program typically includes the following items:

• Axial stiffness in compression and extension, N/mm.
• Lateral (transverse) stiffness, N/mm.
• Angular (bending) stiffness, N·m/degree.
• Torsional stiffness, where contractually specified.
• Hysteresis and linearity of the force–displacement characteristic.
• Stiffness under internal pressure (pressure-adjusted spring rate) where the service condition requires it.
• Ring stiffness and related deflection parameters for thermoplastic corrugated pipes, kN/m².
• Stiffness retention after cyclic displacement (fatigue) testing, to quantify degradation of the spring rate.
• Comparison of measured stiffness with theoretical values calculated per EJMA or design specifications.

Testing Methods

Axial Stiffness Determination

The bellows is mounted between rigid platens on a calibrated universal testing machine, with the axis of the bellows accurately aligned to eliminate eccentric loading. The specimen is first preconditioned by applying several full movement cycles between the specified compression and extension limits to remove residual forming stresses and settle the force–displacement response. The bellows is then displaced at a controlled, quasi-static rate (typically in the range of 1 to 5 mm/min) through the rated axial travel, while force and displacement are continuously recorded.

Axial stiffness is evaluated as the slope of the force–displacement curve, normally taken as a secant between defined displacement limits within the rated movement — for example between 25 % and 75 % of rated travel — with compression and extension values reported separately. Because the response is slightly nonlinear and hysteretic, the reporting convention (initial tangent, secant over rated travel, or averaged cyclic slope) must be stated in the test report. Measured values are compared with the theoretical stiffness computed per EJMA; acceptance tolerances are set by the purchaser or the applicable specification and are commonly on the order of ±15 % to ±25 %, subject also to the maximum allowable anchor loads.

Lateral and Angular Stiffness Determination

For lateral stiffness, one end of the bellows is rigidly fixed and a transverse shear force is applied at the free end through a calibrated fixture, or the assembly is loaded as a guided cantilever. The lateral displacement is measured with displacement transducers at defined positions, and lateral stiffness is calculated as the transverse force divided by the resulting lateral offset.

For angular stiffness, equal and opposite forces are applied through rigid lever arms attached to the end fittings, imposing a pure bending moment about the bellows mid-length. The angular rotation is measured using inclinometers or by geometric derivation from multi-point displacement measurements, and angular stiffness is reported as moment per unit rotation. Multi-convolution bellows are tested with the tie rods or hinges removed or neutralized so that the bellows alone resists the movement.

Ring Stiffness of Corrugated Plastic Pipes

Ring stiffness is measured by the parallel-plate method. A conditioned specimen of standard length is placed between two horizontal parallel plates and compressed at a constant rate while the vertical force and diametric deflection are recorded. The force at 3 % diametric deflection is used to compute the ring stiffness Sn using the standard formula Sn = (0.0186 + 0.025 × Yi/di) × F / (L × Yi), where F is the force, L the specimen length, Yi the deflection, and di the mean inside diameter. If the test is evaluated at 4 % deflection, the result is multiplied by the standard correction factor of 0.75. A minimum of three specimens is tested and the results are reported against the declared SN class.

Equipment and Measurement Assurance

Stiffness determinations are performed on universal testing machines verified in accordance with ISO 7500-1, using calibrated load cells and displacement measurement by LVDT or crosshead transducers with resolution sufficient to define the secant slope accurately. All tests are documented with specimen identification, geometry verification, ambient conditions, loading rate, raw force–displacement data, stiffness calculation method, and comparison against specification limits.

Applications

Bellows stiffness data are used across many industries wherever corrugated flexible elements or corrugated pipes absorb thermal expansion, vibration, or settlement:

• Power generation: main steam, hot reheat, and feedwater expansion joints, where spring rates control anchor and turbine nozzle loads.
• Oil, gas, and petrochemical plants: piping expansion joints, flare and stack bellows, and engine exhaust systems.
• HVAC and district energy: thermal-movement compensation in chilled and heating water networks.
• Vacuum, semiconductor, and analytical equipment: edge-welded bellows where predictable stiffness is critical for actuation and sealing performance.
• Bellows-sealed valves: stem bellows whose axial stiffness affects actuator sizing and seat loading.
• Automotive and rail: exhaust decouplers and flexible couplings requiring tuned lateral and axial stiffness for vibration isolation.
• Infrastructure: HDPE and PVC corrugated drainage pipes, where ring stiffness classes determine trench design and long-term deflection limits under soil and traffic loading.

FAQ

What is the difference between bellows stiffness and ring stiffness?

Bellows stiffness (spring rate) applies to metallic bellows and describes the force or moment needed to produce axial, lateral, or angular movement of the convoluted element. Ring stiffness applies to thermoplastic corrugated pipes and describes resistance to diametric deflection under external loads, determined by the parallel-plate method per ISO 9969, GB/T 9647, or ASTM D2412.

Why is measured axial stiffness compared with the theoretical value?

Theoretical spring rates calculated per EJMA depend on measured wall thickness, convolution geometry, and material modulus, all of which vary within manufacturing tolerances. Comparing the measured value with the calculation verifies that the as-built bellows will not impose excessive anchor loads and that the design margin for fatigue life remains valid.

Does internal pressure affect bellows stiffness?

Yes. Internal pressure can measurably increase the effective stiffness of a bellows, particularly for high-pressure multi-ply designs. When the piping analysis is pressure-sensitive, stiffness may be determined with the bellows pressurized to the design pressure using hydraulic or pneumatic means with appropriate safety controls.

How does multi-ply construction influence stiffness?

For multiple plies of equal thickness, total axial stiffness increases approximately in proportion to the number of plies, while stress per ply decreases, which is why multi-ply bellows are used to combine low spring rate with high pressure capacity.

What specimens are required for a stiffness determination?

For metallic bellows, a representative free-length bellows or production assembly with documented geometry and material certification is required. For ring stiffness testing of corrugated plastic pipes, a minimum of three specimens of standard length, cut from different positions along the pipe and conditioned to the standard atmosphere, are needed.

Can stiffness testing be combined with other tests?

Yes. Stiffness determination is frequently combined with stability (column squirm) testing, burst and proof pressure testing, and cyclic displacement (fatigue) testing, with stiffness re-measured after cycling to quantify spring-rate retention as required by project specifications.