Pavel:

The effect of frequency on fatigue lifetimes can be quite complex. Higher frequencies imply higher strain rates which can lead to higher strengths. However, in metal fatigue, to realize this effect you would need to increase the frequencies by two orders of magnitude or more in order to see much of a difference.

The most prominent effect of frequency in fatigue derives from environmental effects. In essence, at lower frequencies because the crack is open for a longer time period each cycle, corrosion processes (or whatever the active environment effect is, such as metal dissolution or the physi-/chemi-adsorption and ingress of hydrogen is) must have time to occur. Thus, lower frequencies can significantly increase the crack propagation rate per cycle (da/dN), i.e., although the life in terms of time may be longer, the life in terms of cycles may be shorter at the lower frequencies. However, this does depend on the kinetics of the the rate-limiting step of the environmental process compared to the time that fresh metal is exposed when the crack is open. In the fatigue of steels in oxidizing environments, for example, where the primary environmental effect can be the anodic oxidation of the freshly created surface at the crack tip, the crack-propagation rate per cycle at a given stress-intensity range will likely be faster at a 10 Hz frequency than at, say, 100 Hz, because the higher frequency will serve to inhibit the time each cycle for the oxidation to occur. This is in contrast to titanium alloys where the oxidation is very fast; here the crack propagation rates are almost identical at 10 Hz up to 1000 Hz, as the attached article shows.

You ask what equipment you would need to measure such high frequency fatigue. If you performing total life (S/N) tests, I would try and find a resonant machine to do that. If your displacements are not too large, many conventional rotating beam or servo-hydraulic testing machines can possibly get to 300 Hz as well. For crack propagation tests, we were able to get as high as 1000 Hz using a specially designed MTS electro-servo-hydraulic testing machine (as described in the attached paper). If you need to go higher than that, you will need an ultrasonic system where you can cycle at frequencies on the order of 20 kHz (as shown in papers by Stanzl and Bathias).

ROR

Dear Robert O Ritchie, thanks for the detailed explanation. It has got much clearer to me. Maybe that's another story, but maybe you know why the servohydraulic machines are so popular, while resonant machines, so-called Vibrophores, consume far less energy and easier to install?

Pavel:

In my experience, servo-hydraulic machines are far better from the perspective of control. They are readily automated and can be used to perform complicated loading spectra or load-shedding profiles to measure thresholds with ease. They can also be used for any type of mechanical testing, not simply fatigue.

I only used one resonant machine - admittedly this was a while ago! Yes, it didn't consume as much energy and did not require water-cooling, three-phase power, etc., etc., but I found it to be far more difficult to program and control and I couldn't use it for other mechanical functions such as uniaxial tensile and fracture toughness testing. Hence my personal preference has always been for servo-hydraulics.

ROR

Dear Robert O Ritchie,

Massive thanks for sharing your experience.

Hope your area of work goes well.

Kind regards,

Pavel

Dear Pavel,

we recently published an article on this topic that you might find interesting:

Article Frequency-dependent fatigue damage in polycrystalline copper...

Although the focus of the article is on copper, we have found similar frequency dependence of fatigue behaviour in structural steels and therefore believe that a similar mechanism to that described in the article could be observed in aluminium alloys.

Thanks, Ivo Šulák . I will have a look into your paper.

Best wished

Pavel

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

https://www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

www.researchgate.net/publication/235177696_Effect_of_Frequency_on_Fatigue_Crack_Growth_Rate_of_Inconel_718_at_High_Temperature

Dear Pavel,

As you have a plan of doing the Giga cycle fatigue (GCF) on Al alloy, GCF needs more number of cycles, which can be achieved either by resonance based or ultrasonic based fatigue testing machines and hence they would be the better option. However, some difficulties arises due to local heating and high strain rates cause anomaly in the fatigue properties. Therefore, care should be taken during the test in order to get the consistent results. You can refer Prof Stanzle and Dr C Bathias who initiated the GCF concept first.

Next come to the frequency effects are different for different materials, their sensitivities to the specific environment. Generally, uniaxial specimens are used in GCF studies. The fatigue strength defined as when the applied stress becomes plateau with number of cycles. Later, it has been realised that the existence of fatigue limit ( pl refer to H Mugrabi and H J Christ paper) is doubtful. Finally, it has been suggested that the onset of PSBs can be treated as fatigue limit. The formation of PSBs are time dependent. Any time dependent process is governed by kinetics are strongly dependent on frequency, temperature and environment.

Next point is on early strain location and hence lower fatigue limit always may not be true. For example, in stainless steels deformation/plasticity induced transformation enhances the (TRIP effect, please see Dr K Sadananda TRIP paper published just now) local toughening leading to the higher fatigue limit.

I hope this will help you.

with best wishs

Nani Babu

T. Weerasooriya, "Effect of Frequency on Fatigue Crack Growth Rate of Inconel 718 at High Temperature," in Fracture Mechanics: Nineteenth Symposium, ed. T. Cruse (West Conshohocken, PA: ASTM International, 1988), 907-923. https://doi.org/10.1520/STP33112S