EVALUATION OF RESIDUAL STRESSES IN TITANIUM ALLOY WELDED BY TIG USING ULTRASONIC METHOD AND EFFECT OF THE MICROSTRUCTURE ON THE Lcr ELASTIC WAVE

EVALUATION OF RESIDUAL STRESSES IN TITANIUM ALLOY WELDED BY TIG USING ULTRASONIC METHOD AND EFFECT OF

THE MICROSTRUCTURE ON THE Lcr ELASTIC WAVE

* I. Hadji 1, R. Badji 1, S. Gachi 1

Welding and NDT Research Center (CSC), BP 64 CHERAGA - ALGERIA Division of Mechanical and Metallurgical

*Corresponding author: [email protected] Abstract

This study consists to determine the superficial residual stresses on Ti-6Al-4V titanium alloy sheets (e = 2 mm) welded by Tingsten Inert Gaz (TIG) process. The residual stresses can increase the fatigue cracks propagation. We will focus in the method of measurements based on ultrasonic using longitudinal critically refracted (Lcr) waves. This method is based on the acoustoelastic effect, which measure the velocity variation of the elastic waves according to the stress state of the material. This can be achieved through a calibration test to determinate the acoustoelastic coefficient (K).

The results show a tensile stresses in the melted zone (MZ), heat affected zone (HAZ) and compression stresses in base metal (BM). The microstructure effect acts on the acoustoelastic constant K. The correction of t0

corrects the overestimated residual stresses in the (HAZ) and (MZ).

Key words: Titanium alloy, welding, residual stresses, ultrasonic, microstructure.

1. Introduction

Titanium alloys, such as Ti-6Al-4V, are widely used in aerospace, power generation and chemical processing applications owing to their high strength-to-weight ratio and good corrosion resistance in many environments [1]. Welding is an effective manufacturing method for joining components to produce useful components and structures. However, the transient weld thermal cycle often results in the development of residual stresses in the weldment. The presence of the residual stresses can impact subsequent mechanical and corrosion properties of the weldment. They play an important role in the strength and service life of structures.

The high industry request for the stress evaluation techniques encouraged development of several methods like X-ray diffraction, incremental hole drilling, and more recently the ultrasonic waves and the Barkhausen noise methods. Many studies clearly showed that there is no universal or absolute method that gives complete satisfaction in the in-situ non-destructive stress monitoring of the mechanical components. Many parameters such as material, geometry, surface quality, cost, and accuracy of the measurement, etc., must be taken into account in choosing an adequate method.

The ultrasonic method was chosen for stress measurement because it is non-destructive, easy to use, and relatively inexpensive. However, it is rather sensitive to the microstructure effects (grains size [2] and structure [3–4]) and to the operating conditions (temperature [5, 6], coupling [7], etc.). The ultrasonic evaluation of the residual stresses requires separation between the microstructure and the acoustoelastic effects. In the case of welding stresses, this separation was only possible in the MZ and the BM zones. Former works [8–9] showed that significant corrections (100 to 200 MPa), depending on the testing material and the welding process, were obtained while performing separately the calibration tests on the MZ and the BM samples. However, the calibration test was not carried out in the HAZ due to its small width.

The goal of this work was to determine the longitudinal residual stress distribution in a TIG weld on Ti- 6Al-4V and to compare the stress distribution to the microstructures of the various weld zones and the microstructure effect on the acoustoelastic constant.

2. Theory of acoustoélasticité 2.1 Acoustoelastic Effect

The ultrasonic evaluation of the residual stresses is based on the acoustoelastic effect. This effect results in the change of the ultrasonic wave velocity according to the material strain state. Hughes and Kelly [10]

established the formalism describing this phenomenon by using the deformation energy expression developed by Murnaghan [11] to the third order. The velocity variation is expressed as a function of the deformations generated by the applied or residual stresses, the Lame’s elastic constants (l, μ), and the Murnaghan’s elastic constants (m, n, l).

In the case of uniaxial load (_=_and _=_=) and wave propagating in the direction 1, as

 

2

0 11V 2 4( 2 ) 2( 2 )m vu(1 2 / )l

           

2

0 12V u [4u v n( / 2) m(1 v)]

      

 

2

0 13V u 4u v n( / 2) m(1 v)

      

Where ρ0 is the free stress material density, _+₂+₃ is the trace of the strain tensor. Under infinitesimal deformation and considering the linear Hooke’s law, these equations can be written as

1 / ( / _ij0)

ij ij ij

d



Where dσij is the stress variation (MPa) according to the directions (i, j = 1, 2, 3), Vij and V0ij are the wave velocities (i is the propagation direction, j is the polarization direction) in the stressed and unstressed materials, respectively. Kij is the acoustoelastic constants associated with the second and third-order elastic constant (MPa−1). Equation (2) provides the residual stresses when the calibration constants V0ij and Kij are known. In the case of a homogeneous material subjected to a uniaxial load, the longitudinal waves propagating in the same direction as the stress can be simply written as:

(V₁₁V₁₁⁰) /V₁₁⁰ K₁₁₁₁ (3) 2.2 Longitudinal Critically Refracted Wave

In the non-destructive testing community, the longitudinal critically refracted wave is known under name

“subsurface longitudinal wave by Basatakaya and Ermolov (1981) [12], Hereafter, the Lcr terminology will be used to describe this wave, which is refracted near the first critical angle θ1C according to the Snell-Descartes law

Sin 1c = 1/ 2 (4)

Where, VL1 and VL2 are the longitudinal wave velocities for the incidence and the refraction media, respectively. Egle and Bray (1976) established the sensitivity of this wave velocity to the strain on rail steel specimens [13], They showed that the observed high velocity relative change is associated to longitudinal waves propagating in the direction of the applied stress.

2.3 Determination of acoustoelastic constant

The value t⁰ is measured directly from the stress-free samples, the acoustoelastic constant K [equation (3)]

is deduced experimentally from a uniaxial tensile test associated with an ultrasonic measurement. K represents the slope of the relative variation curve of the time-of-flight as described by equation (5).

Where, t and t0 are the time-of-flight measured between the two receivers for stressed and unstressed samples, respectively. σ is the applied stress.

Axial loading was performed with a tensile-testing machine up to 30% of the Ti-6Al-4V elastic limit. The tensile testing machine was controlled in force by steps equivalent to 50 MPa.

2.4 Determination of Residual Stresses

The measurements were parallel to the weld axis. The time-of-flight measurements were performed on the melted zone and the heat affected zone and Base metal every 5 mm.

Coupling liquid was renewed for each measurement in order to reproduce the same test conditions. The values of the residual stresses relating to each weld zone were calculated from the following equations

0 0

1/ [( ) / ]

BM

K

t t

t

   

0 0

1/ [( ) / ]

MZ

K

t t

t

   

3. Experimental Procedures 3.1 Sample Description

The material tested is Ti-6Al-4V titanium alloy sheets (e = 2 mm) welded by Tingsten Inert Gaz (TIG) process. To facilitate the experiments, the welded joint was leveled and the front surfaces were finished with a milling machine in order to become flat.

Figure 1. Micrographics of the various weld microstructures 3.2 Measurement Device

The measurement device, shown in Fig. 2, includes a wave train generator synchronized with a digital oscilloscope having a sampling frequency of 5 GHz.

Figure 2. Schematic representation of the Lcr wave measurement setup

Measurements were performed with a piezoelectric transducer, which is composed of one emitter (E) and two receivers (R1 and R2). Double reception provides the advantage to eliminate the environment effect (temperature, coupling conditions, etc.). The nominal frequency and diameter of the piezoelectric elements are

5 MHz and 10 mm, respectively. The time-of-flight measurements were taken between the two receiver echoes R1 and R2 using the zero crossing method.

3.3 Evaluation of the Calibration Constants

To evaluate the calibration constants (acoustoelastic constant, free stress time-of-flight) corresponding to the BM and MZ, the calibration samples were taken in the direction parallel to the weld Fig2.

Figure 3. Positions of the calibration samples on the original plate

4. Results and Discussion

4.1 Microstructure Effect on the Acoustoelastic Constant

The calibration curves of BM, MZ samples are shown in Fig. 4.a and Fig. 4.b. Linear curves were obtained during the loading and unloading phases without any hysteresis effect, so the acoustoelasticity law could be established. They show that the microstructures affect significantly the material acoustoelastic behavior.

Namely, the acoustoelastic constant of MZ is 2.14 which is 15% different from that of the BM microstructure.

Figure 4. Influence of the microstructure on the acoustoelastic calibration constant

4.2 Microstructure Effect on the Free-Stress Time-of-Flight

Table 1 shows the free-stress time-of-flight measured in the different zones. t0 corresponding to the MZ, which has a solidification structure, is higher (approximately 2%) than that of BM.

Table 1 Free-stress time-off light results

Zone of the weld BM MZ

Free-stress time-of-flight t0 (us) 6,2432338E-6 6,3325527E-6

4.3 Residual Stress Profile in the MZ

The measured residual stresses versus distances from the weld center, before and after microstructure effect correction, are shown in Fig. 5. When using the BM calibration constants to determine the residual stress in the MZ, the result shows an overvaluation of the stress value. The simultaneous correction of t0MZ (σ = 0) and KMZ involves reduction of the erroneous residual stresses of approximately 100 MPa. This result confirms the strong influence of the melted zone microstructure on the calibration constants K and t0.

Figure 5. Ultrasonic residual stress measurement before and after correction of the MZ calibration constants

5. Conclusion

The microstructure effect acts on both the acoustoelastic constant K and the free-stress time-of-flight t0. The correction of t0 in the two weld zones is very important because it corrects the overestimated residual stresses in the HAZ and MZ zones from approximately 20%.

Finally, this paper confirms the potential of the Lcr waves to accurately evaluate the welding residual stresses, if all the sample microstructures are taken into account.

6. References

[1]. Macherauch E, Kloss KH, Proc international conference on residual stresses. Gramish-Partenkirchen (FRG), pp 3–26, (1986).

[2]. Palanchamy P, Joseph A, Jayakumar T, Ultrasonic velocity measurements for estimation of grain size in austenitic stainless steel. NDT E Int 28(3):179–185, (1995).

[3]. Cantrell JH, Salama K, Acoustoelastic characterization of materials. Int Mater Rev 36:125–145, (1991).

[4]. Hakan Gür C, Çam İ, Comparison of magnetic Barkhausen noise and ultrasonic velocity measurements for microstructure evaluation of SAE 1040 and SAE 4140 steels. Materials Charact, 58(5):447 454C, (2007).

[5]. Salama K, Relationship between temperature dependence of ultrasonic velocity and stress. In:

Thompson DO, Chimenti DE (eds) Review of progress in quantitative non-destructive evaluation.

Plenum, New York, pp 1109–1119, (1985).

[6]. Mohbacher H, Schneider E, Goebbels K , Temperature dependence of third-order elastic constants. Proc 9th international conference on experimental mechanics, Aaby Truk, Copenhagen 3:1189–1197, (1990).

[7]. Lhémery A, Calmon P, Chatillon S, Gengembre N, Modeling of ultrasonic fields radiated by contact transducer in a component of irregular surface. Ultrasonics 40:231–236, (2002).

[8]. Hoblos J, Evaluation des contraintes résiduelles par méthode ultrasonore avec prise en compte des effets de la microstructure: application au cas de soudage. PhD Thesis, order number 3456, University of Sciences and Technologies of Lille, France, (2004).

[9]. Qozam H, Bourse G, Robin C, Walaszek H, Bouteille P, Cherfaoui M, Influence parameters on the determination of residual stresses by ultrasonic method: application to the case of pressure equipments.

Eur Symp Press Equip, Paris, pp 501–509, (2007).

[10]. Hughes DS, Kelly JL, Second-order elastic deformation of solids. Phys Rev 5(92):1145–1149, (1953).

[11]. Murnaghan FD, Finite deformation of an elastic solid. Wiley, New York, (1951).

[12]. Basatskaya LV, Ermolov IN, Theoretical study of ultrasonic longitudinal subsurface waves in solid media. Defektoskopiya 7:58–65, (1981).

[13]. Egle DM, Bray DE, Measurement of acoustoelastic and third-order elastic constants for rail steels. J Am Soc Acous 60 (3):741–744, (1976).