Improving fa gue performance of surface-damaged Ti-6Al-4V alloy via ultrasonic surface rolling process

Surface damages caused by markings, like laser marking and mechanical marking are detrimental to the fa gue performance of the components. To improve the fa gue performance of laser marked or mechanical marked Ti-6Al-4V alloy, ultrasonic surface rolling process (USRP) was performed on Ti-6Al-4V alloy prior to marking. The results showed that laser marking induced a porous and loose oxidized layer which severely decreased the fa gue performance of the substrate. And the mechanical marking slightly decreased the fa gue performance of the substrate due to the local stress concentra on at the bo�om of the pits. A�er USRP pretreatment, the detrimental effect of laser marking on the fa gue performance of the substrate was eliminated while the fa gue performance of mechanical marking sample was significantly improved by a factor of 3. That was because the USRP-induced residual compressive stress suppressed the crack early propaga on.


Introduc on
Data matrix (DM) codes have been widely used in the adver sing, internet, electronic cket business and industry due to its small area, large amounts of informa on, high reliability, high error-correc ng ability and high security [1][2][3]. There are many methods, like ink-marking, electro-chemical methods, mechanical marking and laser marking to mark DM codes to the products [1]. Among them, laser marking produces permanent, high contrast codes rapidly with rela vely low cost [3]. However, the laser marking causes a large surface damage on the substrate which is not conducive to its fa gue proper es. Mechanical marking could also produce a permanent code with rela vely small surface damage on the substrate. Titanium alloys are a�rac ve due to their high specific strength, excellent corrosion resistance and biocompa bility [4]. However, tanium alloys are sensi ve to surface notch effect. Thus, it is necessary to inves gate the fa gue performance of surface-damaged tanium alloy.
Ultrasonic surface rolling process (USRP), an emerging surface mechanical treatment, could induce a deep-distributed compressive residual stress layer and an excellent surface integrity [5,6]. It is reported that USRP could significantly improve the MATEC Web of Conferences 321, 09003 (2020) https://doi.org/10.1051/matecconf/202032109003 The 14 th World Conference on Titanium fa gue performance of materials [5,6]. Therefore, it is proposed that USRP treatment could be performed prior to laser marking or mechanical marking to improve the fa gue performance.
In the present study, Ti-6Al-4V alloy, the most widely used tanium alloy all over the world, is chosen as the substrate. The fa gue performance of the base material (BM), laser marked sample (LM), USRP-laser marked sample (USRP-LM), mechanical marked sample (MM) and USRP-mechanical marked sample (USRP-MM) is inves gated.

Experimental material
The material used in this study was received as a bar of Ti-6Al-4V alloy supplied by Baotai Co. Ltd. in China. The nominal chemical composi on, heat treatment condi on and the mechanical proper es of the Ti-6Al-4V alloy had been elaborated in Ref [7]. The dog-bone shaped fa gue specimens (the working sec on diameter is Ф6 mm) were machined from a Ф15 mm bars for conduc ng the fa gue tests.

Marking methods and USRP parameters
A YLP-D10 op cal fiber laser marking machine and a TMP6100 mechanical marking machine were used in the experiment.
The marking speed of 400 mm/s, laser power of 68 W, scanning pitch of 0.02 mm and module indent of 0.056 mm were used in laser marking. In mechanical marking, the marking speed of 50%, the pressure of 3 MPa and the distance of 6 mm were used.
The specimens were subjected to USRP under a sta c force of 900N using a rota on speed of 200 rev/min and the feeding rate is 0.10 mm/rev. A scrollable rolling WC/Co ball with a hardness of 80 HRC, surface roughness of Ra 0.1 μm and radius of 14mm equipped with an ultrasonic apparatus was used for process. The material was subjected to an ultrasonic vibra on frequency and ultrasonic vibra on amplitude of 20 kHz and 20 μm, respec vely.

Material characteriza on
The surface morphologies and fracture morphologies of all the samples were observed using scanning electron microscopy (SEM) (JEOL JSM-6390). The element content was detected by energy dispersive spectrum (EDS) instrument a�ached on the SEM device.

Fa gue test
To evaluate the effect of different surface damages on the fa gue life of Ti-6Al-4V alloy, rota ng bending fa gue tests were performed at a PQ-6 type rota ng bending fa gue tester at room temperature. A�er laser treatment, a porous and loose layer with a depth of 40~60 μm was formed on the tanium alloy surface. And the treated area had been oxidized via the EDS analysis. The porous and loose layer were easy to crack under the fa gue loading [8].
Many literatures showed that the porous and loose surface layer was harmful to its mechanical proper es [9]. A�er mechanical marking, many pits with a depth of 30~35 μm were formed on the tanium alloy. And the bo�om of the pits was blunt.    The DM codes generally have a deleterious effect on the substrate due to surface damages caused by them. However, it is well known that surface mechanical treatments, like shot peening, surface mechanical rolling treatment and USRP, significantly improved the fa gue performance of metal materials [6,10,11]. Therefore, the sample was treated by USRP prior to marking to improve the fa gue performance of laser/mechanical marked tanium alloy. The results showed that USRP pretreatment did improve the fa gue performance of laser/mechanical marked tanium alloy. The fa gue life of LM sample was improved similar to that of the substrate via USRP pretreatment while the fa gue life of MM sample was improved by a factor of 3 via USRP pretreatment. This was mainly a�ributed to the compressive residual stress induced by USRP pretreatment. Our previous study [6] showed that USRP treatment could induce a large value compressive residual stress layer with a depth of ~530 μm at the tanium alloy surface layer. And the maximum value is 930 MPa at the depth of ~50 μm. It is well known that compressive residual stress could retard the fa gue crack ini a on and fa gue crack early propaga on [12]. However, the fracture morphologies of USRP-LM sample and USRP-MM sample showed that the fa gue crack ini a on sites of the two samples were the same as that of LM treated sample and MM treated sample, which was ini ated on the surface. This was different from the 5 MATEC Web of Conferences 321, 09003 (2020) https://doi.org/10.1051/matecconf/202032109003 The 14 th World Conference on Titanium surface mechanical treated samples with no surface damage, the fa gue crack ini a on sites of which were generally located at the material interior [6]. This illustrated that the compressive residual stress had li�le influence on the fa gue crack ini a on sites but a big influence on the fa gue crack propaga on. The compressive residual stress could not only decrease the tensile stress at the crack p but also close up the cracks in the propaga on early stage [13,14]. Thereby, the compressive residual stress slowed down the crack propaga on rate. the laser marked DM code should avoid the complete oxidized line of the code to improve the fa gue performance. In addi on, the cracks of the mechanical marked sample were ini ated at the pits. It was suggested that the process parameters of the mechanical marking should be improved to obtain rounder pits with less local stress concentra on.

Conclusion
(1) The laser marking caused a porous and loose oxidized layer with a depth of 40-60 μm on the tanium alloy surface while the mechanical marking caused many pits with a depth of 30-35 μm on the tanium alloy surface.
(2) Compared with the substrate, laser marking severely reduced the fa gue performance of the tanium alloy. This was a�ributed to the fact that laser marking produced a notch surface with loose oxide layer that was easy to ini ate cracks. In addi on, mechanical marking reduced the fa gue performance slightly for the bo�om of the iden fica on pits was blunt and therefore the local stress concentra on caused by the pits was slight.
(3) USRP pretreatment significantly improved the fa gue life of mechanical marked sample while USRP pretreatment improved the fa gue life of laser marked sample similar to that of the substrate owing to that the large value and deepdistributed compressive residual stress retarded the fa gue crack early propaga on.