Fracture failure analysis of baseplates in a fluidic amplifier made of WC-11 Co cemented carbide

A fluidic amplifier is a crucial automatic control component in a liquid jet hammer used to drill hard formations in the oil industry. This study aims to determine the true causes of the fracture failure of baseplates in a fluidic amplifier made of WC-11Co cemented carbide in a very short period of time. Computational fluid dynamics (CFD), theoretical estimation, and finite element analysis (FEA) were employed to analyze the effect of static and dynamic loads on the strength of the baseplates. Fractographic, metallographic, and processing defect analyses were also carried out. The FEA results showed that the static and dynamic loads caused stress concentrations at the actual fracture locations, and the effect of working loads on material strength was allowable and safe. Fracture surfaces exhibited typical characteristics of a brittle fracture. The metallographic analysis revealed that a specific amount of brittle eta-phase (η-phase) was present in the material. The microstructure of the processing cutting zone was inspected and the results revealed that some voids, pores, and microcracks were formed on the processing surface. The manufacturing and processing defects resulted in low stress fracture failure of the baseplates.


INTRODUCTION
n an oil drilling field, drilling hard formations is an arduous problem that demands prompt solution all the time, especially in deep wells.As deep-oil resource exploration continues to increase, the depths of oil wells have become increasingly deeper and the rocks at such great depths have become harder to drill.The hydraulic hammer was considered as one of the best tools to drill hard formations for many years [1][2][3].The application of conventional rotary drilling to drill hard formations have several weaknesses such as high cost, low rate of penetration (ROP), and short service life of drilling tools [4][5][6][7][8][9].The liquid jet hammer technique is a highly efficient method that overcomes the disadvantages of conventional drilling methods for hard formations.The liquid jet hammer is a down-the-hole tool that is used to exert an impacting force on the drill bit [4].It has been widely applied in exploration core drilling since its invention in the 1970s and was crucial in drilling the No. 1 hole of the Project of China Continental Science Drilling [5][6][7].Exploration core drilling efficiency can be improved by more than 30%, higher core recovery rates can be obtained, and straighter holes can be drilled when a liquid jet hammer is used [4,5].In oil and gas drilling, a number of field tests have proven that liquid jet hammer could increase drilling efficiency by more than 120% in hard formations [8][9][10].As shown in Fig. 1, the YSC178 liquid jet hammer [5,11] is placed at an oil drilling platform of the Zhongyuan oil field and is connected at the bottom of the drill pipes.

DESCRIPTION OF THE FLUIDIC AMPLIFIER AND THE LIQUID JET HAMMER
bistable fluidic amplifier comprises two baseplates, two lid plates, and a splitter.These components are connected to one another by pins and bolts, as shown in Fig. 2. The supply jet fluids that enter the fluidic amplifier have several flow characteristics, such as attachment, switching, and vortex.The supply jet switching mechanism is a feedback loop that is designed between the control nozzle and the output channel on each side, which is relatively similar to the feedback loop in a fluidic oscillator for flow metering or flow separation control [4].channel (f).The high-pressure jet fluids push the impact body to return to its original position.The fluids in the upper chamber of the cylinder are extruded into the annular space between the hammer and outer sleeve through the vent (c).When the piston moves to the stopping end of the cylinder, the fluid pressure in the lower chamber of the cylinder increases.The pressure signal propagates along the feedback loop and reaches the control nozzle (b), which signals the jet fluids attached to the right sidewall to switch to the left sidewall [4,5].The bistable fluidic amplifier controls the high-pressure drilling fluids to enter alternately in the upper and lower chambers, which causes the reciprocating motion of the impacting body.Impact force is generated periodically and impact energy is transferred to the anvil and tricone roller bit.

DESCRIPTION OF FAILURE SCENARIO
he fluidic amplifier is an automatic control device that uses several fluid properties [12] and is regarded as the most critical part in a liquid jet hammer.Therefore, the service life of a fluidic amplifier has an important effect on the normal work of a liquid jet hammer.The work condition of a fluidic amplifier is extremely complex because of high-velocity drilling fluids that contain abrasive particles.In previous oil-drilling applications, the velocity of the drilling fluids that enter the supply nozzle reaches 60 m/s to 100 m/s, and the failure of a fluidic amplifier is mainly due to abrasive erosion [13,14].High-velocity solid particles in drilling fluids wear away the inner structures of a fluidic amplifier, and significant erosion pits can be seen on the inner surface of a fluidic amplifier.Abrasive erosion leads to the failure of the fluidic amplifier such that the service time of a fluidic amplifier ranges from a few hours to a dozen hours.The fluidic amplifier in this study was made of WC-11Co cemented carbide, as shown in Fig. 4 (a), with an expected service life of 120 h in complex conditions.However, the accident occurred that the liquid jet hammer only worked within several seconds in the test.The fluidic amplifier from the YSC178 liquid jet hammer was taken out, and three fracture locations were found at the baseplates of the fluidic amplifier, as shown in Fig. 4(b).The appearance of the fractured section was smooth and no evident plastic deformation, such as elongation or bending, were observed.No signs of abrasive erosion were observed on the surface of the sidewalls and the supply nozzle.This form of failure was noticeably different from those in previous tests [13,14].This failure was evidently a brittle fracture, but not abrasive erosion.WC-11Co cemented carbide is a typical brittle composite material; fracture failure exhibits no distinct plastic deformation.The yield strength of this material is not easily determined by experimental tests.A sudden fracture failure may occur in T WC-11Co cemented carbide if the actual local stress exceeds the material strength limit.In addition to different unexpected failure scenarios, the manufactured material and loading situation of the fluidic amplifier in this study are different from those in previous tests.(1) New material.WC-11Co cemented carbide was used to manufacture a fluidic amplifier instead of 35CrMo alloy steel in the previous test to reduce abrasive erosion.This material exhibits several good material properties, such as high hardness, high strength, good wearing resistance, and great corrosion resistance [14].
(2) No disc springs.A set of disc springs used in the previous test was installed at the top of the fluidic amplifier.These disc springs were adjusted to provide a small fitting allowance among the components inside a liquid jet hammer; in this way, sealing problem could be solved and cylinder deformation could be avoided.However, the disc springs with an unreasonable design and unpredictable fitting allowance can lead to more serious leakage and abrasive erosion.Therefore, the disc springs in this test were removed from a liquid jet hammer [13,14].

NUMERICAL SIMULATIONS FOR STRUCTURE STRENGTH ANALYSIS OF THE FLUIDIC AMPLIFIER
n this study, the simplified physical process explains that before the liquid jet hammer has applied work on the drilling fluid pressure and the axial static pressure caused by threaded connection have generated pre-stress on the fluidic amplifier, the impact force caused by the backward stroke of the impacting body is exerted at the bottom of the fluidic amplifier.The force analysis diagram of the fluidic amplifier is shown in Fig. 5.However, the effect of the static and dynamic loads on the material strength of the fluidic amplifier remains unknown.The "implicit-explicit sequence solving" analysis method [15] is employed to investigate the reason for fracture failure.The following is the procedure of the numerical analysis: (1) the drilling fluid pressure on the sidewalls of the baseplate and the terminal velocity of the backward stroke of the impacting body can be calculated using the CFD method; (2) the axial static pressure on the fluidic amplifier can be theoretically estimated; (3) the calculated drilling fluid and axial static pressures are set at the boundary conditions in the implicit static calculation analysis, and pre-stress static results could be obtained by an implicit solver; (4) based on the implicit results, the terminal velocity of the backward stroke of the impacting body is set to an initial condition in the explicit solver; (5) the dynamic load results can be obtained using the explicit solver of ANSYS/LS-DYNA.The effect of the static and dynamic loads on the material strength of the fluidic amplifier can be determined by comparing the first-principle and equivalent stresses to the strength limit.

Computational domain and mesh model
his study combined the simulation of the fluid and the motion of the impacting body by using the dynamic mesh technique and a user-defined function(UDF) [16][17][18][19][20][21].The computational domain of the fluid field in the YSC178 liquid jet hammer includes the fluid field of a fluidic amplifier, two side passages connected to the vents of the fluidic amplifier, the upper and lower chambers of the cylinder, as well as the connecting passages between the fluidic amplifier and the cylinder.The computational domains were meshed using the Altair Hypermesh software.The mesh model was used as the starting volume mesh for the dynamic mesh modeling technique.The mesh of the computational domains was shown in Fig. 6.In this mesh model, 1636 out of 37713 elements were pentahedron, and the rest were hexahedral.The mesh dependence test was performed for the mesh model of the YSC178 liquid jet hammer.Three sets of meshes were tested.As shown in Tab. 1, the maximum difference between the obtained results is less than 5.2%.The coarse mesh can also provide sufficient mesh independency.Considering the computational time cost versus the model accuracy, computations were performed using the medium mesh.

T Initial and boundary conditions of CFD
The computation had two steps: the steady and unsteady simulations.The result of the steady computation provided the initial conditions for the unsteady simulation.
In the steady simulation step, a velocity inlet boundary condition was set at the supply nozzle inlet of the fluidic amplifier (Fig. 6).At the experimental test, the pump flow rate was 0.02 m 3 /s, and the velocity inlet boundary condition was set at 71.4 m/s.Two vents and side passages provided two outlets in the computational domain.Two outlet faces, namely, Bb and Bu, were defined as the pressure outlet boundary conditions.Bb and Bu were the outlet faces of the side passage connected to the lower and upper chambers of the cylinder, respectively.At Bb, the pressure outlet boundary condition was set at atmospheric pressure.At Bu, the pressure outlet boundary condition was set at least two times higher than the atmospheric pressure to enforce the main jet attached to the sidewall corresponding to the lower chamber.In this manner, the upward movement of the impacting body was initiated.
In the unsteady simulation step, the velocity inlet boundary condition was the same as that in the static simulation.However, the pressure outlet boundary conditions were set to atmospheric pressure at both Bb and Bu.Thus, the possibility is the same as that of the supply jet attached to either sidewall when the impacting body reached the stopping end of the cylinder.

CFD results
The pressure distribution (Fig. 7) illustrates that the fluid pressure on the sidewalls is about 0.7 MPa when the flow rate in the test was 0.02m 3 /s.The velocity distribution (Fig. 8) illustrates that the terminal velocity of the backward stroke of the impacting body is about 1.85 m/s.The fluid pressure on the sidewalls is one of the boundary conditions in the "IMPLICIT ANALYSIS" section.The terminal velocity of the backward stroke of the impacting body is one of the initial conditions in the "EXPLICIT ANALYSIS" section.28 9107 In the section of Implicit analysis, the axial static pressure at the top of the fluidic amplifier was 28 MPa.

STRUCTURE STRENGTH ANALYSIS USING FEA
tructural implicit-to-explicit sequential solutions can be used for a broad range of engineering problems, wherein the initial static stress state of the structure affects its dynamic response in ANSYS LS-DYNA [15,23].
In this implicit-to-explicit sequential solution, a calculated drilling fluid pressure of 0.7 MPa and an axial static pressure 28 MPa are set to be preloads in the ANSYS implicit analysis.The nodes of any element that are only used in the explicit analysis were completely constrained.The information of nodes and elements obtained from the ANSYS implicit solver are written in the ANSYS LS-DYNA dynamic relaxation file.The explicit analysis can be conducted after setting the initial velocity at 1.85 m/s and other calculated conditions.The explicit solver uses the displacement results stored in the relaxation file for the stress initialization of the model.The explicit analysis begins at time zero with a stable preloaded structure.Then the explicit analysis can be solved and completed.

Mesh model
The simplified geometry models of the fluidic amplifier, the cylinder, and the impacting body are built.Material and element properties corresponding to each component are shown in Tab. 3. The material attributes of 35CrMo were assigned to the mesh models of the cylinder and the impacting body.The material attributes of WC-11Co were assigned to the mesh model of the fluidic amplifier.In the implicit analysis, the element type and material model of the fluidic amplifier were defined as SOLID 185 and the linear elastic model, respectively.In the explicit analysis, the element type and material model of the fluidic amplifier were defined as SOLID 164 and the bilinear isotropic hardening model, respectively.The material models of the cylinder and the impacting body were defined as the linear elastic model and the rigid body, respectively.The element type of the cylinder and the impacting body was the same as that of the fluidic amplifier.
A short, solid cylinder was used to replace the actual cylinder for simplification, and the thickness of cylinder was defined as the actual value.The mesh model contained the fluidic amplifier, the cylinder and the impacting body.Some weak areas were meshed finely, such as those in control nozzles, sidewalls, and vents.The mesh model is shown in Fig. 9.In this mesh model, 2473 out of 47455 elements were pentahedron, whereas the rest were hexahedral.The mesh dependence test was performed for the mesh model of FEA.Three sets of meshes were tested.As shown in Tab. 4, the maximum difference between the obtained results is less than 4.5%.The coarse mesh was observed to provide a sufficient mesh independency.Thus, the computations were performed using the medium mesh.a The percentage difference between the coarse and fine mesh.
Table 4: The details of the mesh dependence test for the mesh model of FEA.

Implicit analysis
An axial static pressure of 28 MPa was applied on the top of the fluidic amplifier, whereas a fluid pressure of 0.7 MPa was applied on the sidewalls.After defining the necessary boundary conditions, the implicit analysis can be conducted.The equivalent stress distribution of the implicit analysis is shown in Fig. 10.The equivalent stress concentration of 251 MPa appeared at the location between the control nozzle and the vent.The maximum equivalent stress was too small to cause the fracture failure.
All results that contained the node, element, constraint, and boundary information were written in the relaxation file, which were prepared for the explicit analysis.

Explicit analysis
Boundary conditions, initial conditions, and contact parameters were redefined on the basis of the actual situation after the relaxation file was initialized.A terminal velocity of 1.85 m/s, which was calculated using the CFD method, was applied to define the initial velocity of the impacting body.An impacting body mass of 63 kg was used to define the inertia options.Although stress variation was complex when the baseplates were subjected to a dynamic impact load, the maximum stress was less than the strength limit of approximately 2100 MPa.In a typical operation of a liquid jet hammer, the fluidic amplifier was in a safe working state.Thus, fracture failure within a very short period of time was impossible to occur in an intact fluidic amplifier under normal working conditions.However, the FEA results revealed that the static and dynamic loads generated stress concentration at the actual facture locations.According to the structure strength analysis by FEA, low strength and abnormal loads can be ruled out as the causes of failure.However, once the material showed minor defects, a low stress brittle fracture can occur within a short period of time.This assumption is consistent with the failure scenario in Fig. 4. Therefore, some tests for the failed material were conducted in the following contents.12(e) and 12(f).These pores can easily cause the discontinuity among each phase and form the crack origins.These defects produced stress concentration and reduced material strength.All of the fracture characteristics were typical brittle fractographic observations.The bright edges and surfaces in Fig. 12 were the carbon-deficient features, in which the carbon-deficient phase can increase material brittleness.Therefore, the microstructure of the three fracture locations of the baseplates exhibited a typical brittle fracture.

S Metallographic analysis
Three metallographic sections, which were respectively perpendicular to the fracture surfaces A, B, and C in Fig. 4, were prepared for the observation.After a color developing agent was applied, some non-uniform distributed eta-phases appeared on the polished surfaces (Fig. 13).The dark regions in the metallographs were the eta-phase regions.Eta-phase (η-phase) is a complex carbide compound of W and Co. η-phase has a formula of x y z W Co C .Eta phase is also a carbondeficient form of the cemented carbide with hard and brittle performance.This material is generally considered harmful to the material performance of cemented carbide.In WC-11Co carbide sintering, lower carbon content generated higher eta-phase content and more pores.These carbon-deficient phases can significantly reduce material strength.

Processing defect analysis
Given the complex shapes of the baseplates and high hardness of WC-11Co cemented carbide, conventional machining operations cannot be achieved.The baseplates of a fluidic amplifier were processed by the wire electrical discharge machining (WEDM) cutting method.WEDM is a non-traditional, thermoelectric process, which erodes material from the work piece by a series of discrete sparks between a work electrode and a tool electrode immersed in a liquid dielectric medium.These electrical discharges melt and vaporize minute amounts of the work materials, which are then ejected and flushed by using a dielectric medium [24,25].However, the microcracks on the cemented carbide easily developed during machining.The crack initiation and voids of the WEDM cutting zones adjacent to the fracture surfaces were detected using SEM (Fig. 14).Many voids, pores, and microcracks were observed by SEM.These defects can reduce material strength and cause total failure.A large thermal gradient between the cutting surface and internal material structure during WEDM can generate microcracks and thermal residual stresses.

DISCUSSIONS AND CONCLUSIONS
series of numerical simulations and tests were performed.The main conclusions are summarized as follows: 1.The static and dynamic loads generated stress concentration and great stress variations at the actual fracture locations using CFD and FEA under normal working states of a liquid jet hammer.However, the maximum stresses were less than the material strength limit of WC-11Co cemented carbide.The effect of static and dynamic loads on the material strength of the baseplates was allowable and safe.A 2. The SEM microstructures of the fracture surfaces revealed that clear characteristics of a brittle fracture can be observed.Such characteristics were transgranular and intergranular cleavage fracture, pores, secondary cracks, and carbondeficient phase.The metallographic analysis revealed that the presence of a specific amount of eta-phase can increase material brittleness.The processing defects of the WEDM cutting method were observed under SEM.Many voids, pores, and microcracks were found on the cutting surfaces at the vicinity of the fracture locations.These defects can reduce material strength and cause total failure.3. The results of numerical simulations and experimental tests indicated that the baseplate fracture accident was the brittle fracture failure with low stress.Fracture failure was attributed to the manufacturing defects in WC-11Co cemented carbide sintering and the processing defects of the WEDM cutting method.

IFigure 1 :
Figure 1: Test platform of a liquid jet hammer at the Zhongyuan oil field.

Figure 2 :
Figure 2: Configuration of the bistable fluidic amplifier in a liquid jet hammer.(a) Whole amplifier; (b) Amplifier removed a lid plate; (c) main fluid domain of an amplifierThe liquid jet hammer has several major parts, namely, fluidic amplifier, cylinder, piston, mass block, and anvil.A bistable fluidic amplifier is employed to regulate the actuator of a liquid jet hammer.The actuator comprises a cylinder, a piston, and a mass block.The piston and the mass block were linked using a cone connection to form an impacting body.The working principle of a liquid jet hammer is shown in Fig.3.The drilling fluids enter the liquid jet hammer through the supply nozzle on the fluidic amplifier and are turned into high-pressure jet fluids.The jet fluids randomly attach to the left or right sidewall.(1) Downward stroke of the impacting body: The jet fluids attach to the left sidewall and enter the upper chamber of the cylinder through the output channel (e).The high-pressure fluids push the impact body to strike the anvil.The fluids at the lower chamber of the cylinder are extruded into the annular space between the hammer and outer sleeve through the vent (d).When the piston moves to the stopping end of the cylinder, the fluid pressure in the upper chamber of the cylinder increases.This pressure signal propagates along the feedback loop and reaches the control nozzle (a), which signals the jet fluids attached to the left sidewall to switch to the right sidewall.(2) Backward stroke of the impacting body: The jet fluids attach to the right sidewall and enter the lower chamber of the cylinder through the output

Figure 3 :
Figure 3: Working principle of a liquid jet hammer and a fluidic amplifier.

Figure 4 :
Figure 4: Intact and damaged baseplates of the fluidic amplifier.(a) Intact fluidic amplifier; (b) three fracture locations on the baseplates.

Figure 6 :
Figure 6: Computational domain and boundary conditions.

a
The approximate fluid pressure on the sidewalls of the fluidic amplifier.bThe terminal velocity of backward stroke of the impacting body.cThe percentage difference between the coarse and fine mesh.

Figure 7 :
Figure 7: Pressure distribution when the flow rate was 0.02 m 3 /s.Figure 8: Velocity distribution when the impacting body reached the stopping end.

Figure 9 :
Figure 9: Mesh model of FEA Number of elements

Figure 10 :
Figure 10: The equivalent stress distribution of the implicit analysis.

Figure 11 :
Figure 11: Typical stress distributions of the explicit analysis.(a) and (b) are first-principle stress distribution; (c) and (d) are equivalent stress distribution.The typical results of the explicit analysis are shown in Figs.11(a-d).Figs.11(a and b) and 11(c and d) show the firstprinciple and equivalent stress distributions at different times, respectively.In Fig.11(a), the maximum first-principle canning electron microscopy (SEM) was performed to investigate the microstructure of the fracture locations A, B, and C on the baseplates.The smooth fracture morphology and the absence of distinct ductile fracture features, such as ductile dimples, are shown in Figs.12(a) and 12(c).Many pores were visible.Figs.12(b-f) show some transgranular and intergranular cleavage fracture surfaces."River line" marking patterns, fine pores, and secondary cracks are shown in Figs.

Figure 12 :
Figure 12: SEM micrographs of the fracture locations A, B, C on the baseplates in Fig.4(b).(a) and (c) are the fractographs of location A and B, respectively; (b) and (d) are magnified figures of (a) and (c), respectively; (e) and (f) are the fractographs of location C.

Figure 13 :
Figure 13: Metallographic sections of the fracture surfaces.(a), (b) and (c) are the metallographs of the fracture surfaces A, B and C in Fig.4, respectively.

Figure 14 :
Figure 14: SEM views of cutting zones adjacent to fracture surfaces.

Table 1 :
The details of the mesh dependence test for the YSC178 liquid jet hammer.

Table 2 :
The main parameters of API 5 1/2 IF thread.

Table 3 :
Material and element properties of the components