New PBF Ti-6Al-4V Study Shows Test Frequency Can Select the Defect That Fails

A study made available online on July 1, 2026 in the International Journal of Fatigue shows why a PBF Ti-6Al-4V fatigue result cannot be transferred by cycle count alone. The researchers tested laser powder bed fusion material at 10, 100, 1,000 and 20,000 Hz and found a strong apparent frequency effect on fatigue life. Their fracture-mechanics analysis then showed that the underlying defect tolerance was largely frequency-independent.

The distinction matters. Higher frequency changed which defects were sampled and which crack-initiation mechanism controlled failure. It did not simply make the titanium alloy intrinsically more resistant to damage.

For buyers, engineering teams and suppliers, that is a test-to-service warning. An ultrasonic result can efficiently reach the very-high-cycle regime, but its smaller specimen, stressed volume, control method and exposure time may present a different competition among surface defects, lack-of-fusion flaws and internal keyhole pores than the production part sees.

One Frequency Axis Contained Several Test Changes

The study used identical PBF-LB/M Ti-6Al-4V batches produced in three conditions: recommended process parameters, parameters promoting lack-of-fusion defects and parameters promoting keyhole defects. All were stress relieved at 704°C for 1 hour under argon. The team characterized defects with X-ray computed tomography and electron microscopy before comparing constant-amplitude fatigue results.

Tests at 10, 100 and 1,000 Hz were force-controlled; the 20,000 Hz ultrasonic test used displacement control and a smaller resonant specimen. Run-out limits also expanded with frequency, from 10^7 cycles at 10 and 100 Hz to 10^8 at 1,000 Hz and 10^9 at 20,000 Hz. Cooling and intermittent loading were used at the higher frequencies to manage specimen heating.

At fixed stress amplitude, fatigue life increased markedly from 10 to 20,000 Hz. Across the broader S–N data, crack initiation tended to move from surface or near-surface locations at higher stress toward internal defects at lower stress and longer life. The exact pattern varied with process condition. The keyhole condition, for example, showed a strong preference for volume-initiated failure.

Those observations could be misread as a material improvement caused by frequency. The authors instead applied a Murakami-style fracture-mechanics analysis using defect size and effective stress intensity. The frequency separation largely disappeared in the ΔK–N representation. Their interpretation was that frequency mainly changed the governing crack-initiation site and effective defect severity, rather than fundamentally changing the material’s intrinsic damage tolerance (open-access paper).

The Test Can Select A Different Fatal Defect

This is the industry mechanism: fatigue testing is a competition among available initiation sites. Stress amplitude, environment, stressed surface, stressed volume and time per cycle determine which candidate wins. When the specimen or method changes, the population presented to the test also changes.

At lower frequencies and higher stresses, surface-connected flaws receive more time for oxidation, adsorption-assisted slip and local plastic processes. At higher frequencies and lower stresses, those time-dependent effects can be suppressed while internal defects become relatively more important. A smaller ultrasonic specimen also has less highly stressed volume in which to encounter a large internal flaw.

The paper does not say frequency is irrelevant. It says the apparent life difference needs decomposition. A test at 20,000 Hz may reach 10^9 cycles in a practical time, but it cannot automatically reproduce the defect population, load control, thermal behavior, humidity, dwell, multiaxiality or stressed volume of an aerospace bracket, medical component or rotating industrial part.

That boundary protects against two opposite mistakes. A buyer should not reject ultrasonic fatigue data merely because service frequency is lower; the data can reveal long-life crack initiation and defect sensitivity. A supplier should not use a high-frequency run-out as direct proof of component life without a justified bridge to the real part and load spectrum.

A Six-Axis Test-To-Service Comparison

Before fatigue data supports a released PBF titanium component, six axes should be compared explicitly.

The framework asks for equivalence, not identical test hardware. A defensible bridge can combine rapid ultrasonic screening, lower-frequency confirmation, defect characterization, analysis and representative component testing. The mix depends on consequence of failure and the governing specification.

What Buyers And Suppliers Can Use Now

The paper gives process teams a more useful target than “minimize porosity.” Lack-of-fusion and keyhole conditions create different defect morphologies and locations. Mean porosity can hide the flaw that dominates the actual stress field. Characterization should therefore connect the manufacturing condition to defect shape, position and effective fracture severity.

For a buyer combining special titanium alloy requirements with final titanium CNC machining, the surface route belongs in the fatigue file. Machining may remove an as-built surface population and introduce a new roughness, residual-stress or geometric condition. The tested specimen and delivered component should not silently cross that boundary.

Suppliers can also separate development evidence from lot release. A broad frequency program can establish a material and process data basis. Recurring production still needs controls that keep powder, build parameters, heat treatment, surface state and inspection inside that basis. A machine, parameter, heat-treatment, specimen-orientation or finishing change can alter the defect competition even when nominal alloy and density remain unchanged.

The restrained conclusion is not that one frequency is correct. It is that frequency is one coordinate in a larger evidence system. The July study shows that PBF Ti-6Al-4V fatigue life can appear frequency-dependent because the test selects a different fatal defect. Buyers should accept a life claim only after the defect population, loading method, stressed volume and service environment have been made comparable.

Industry FAQ

What did the 2026 PBF Ti-6Al-4V fatigue study compare?

It compared tests at 10, 100, 1,000 and 20,000 Hz across recommended, lack-of-fusion and keyhole process conditions after the same stress-relief treatment.

Did higher frequency change the alloy’s intrinsic defect tolerance?

The fracture-mechanics comparison largely removed the frequency separation, indicating that frequency mainly changed which defects became critical rather than fundamentally changing intrinsic damage tolerance.

Why can ultrasonic fatigue results differ from lower-frequency tests?

Frequency was coupled with stress amplitude, specimen size, stressed volume, control mode and environmental exposure time, so the tested defect population and crack-initiation site changed.

What should a buyer compare before using fatigue data?

Compare material and process condition, defect type and location, stress and load ratio, frequency and control mode, specimen geometry and stressed volume, environment, run-out definition and service load spectrum.