Surface and Coatings Technology 133᎐134 Ž2000. 572᎐582

Influence of a commercial electroless Ni᎐P deposit on the fatigue properties of a notched and unnotched SAE 4140 steel A. Pertuz a , J.A. Berrıos ´ b, E.S. Puchi Cabrerab,U a

School of Mechanical Engineering, Faculty of Engineering, Uni¨ ersidad Central de Venezuela, Apartado Postal 47885, Los Chaguaramos, Caracas 1045, Venezuela b School of Metallurgical Engineering and Materials Science, Faculty of Engineering, Uni¨ ersidad Central de Venezuela, Apartado Postal 47885, Los Chaguaramos, Caracas 1045, Venezuela

Abstract The effect of a commercial electroless Ni᎐P deposit on the fatigue properties of an SAE 4140 steel in the quenched and tempered condition, has been investigated when the substrate is in a notched and an unnotched condition. The application of such a coating to the substrate gives rise to a significant reduction of the fatigue life in comparison with the uncoated samples. The coated specimens in the as-deposited condition showed a reduction of approximately 88%, whereas in the notched uncoated and coated conditions it was of approximately 94᎐95%. The decrease in fatigue properties in the unnotched coated samples is comparable to that reported for the notched uncoated specimens. It has been observed that the dominant crack responsible for the fracture process is nucleated at the root of the notch, regardless of the presence of the deposit. A rough estimate of the fracture toughness of the material has been determined from the results of the experiments conducted with the notched uncoated specimens, together with the values of the critical crack length at different alternating stresses. The material constants involved in the Paris relationship, for the description of the fatigue crack growth rate as a function of the stress intensity factor at the crack tip, were also determined from the results obtained with the notched uncoated samples. 䊚 2000 Elsevier Science B.V. All rights reserved. Keywords: Electroless nickel; Fatigue; 4140 steel

1. Introduction Electroless Ni᎐P ŽEN. deposits have been reported to diminish severely the fatigue properties of highstrength steels w1᎐4x. Among the most recent studies regarding this subject, Wu et al. w1x reported a reduction in the fatigue limit of a 30CrMo steel Ž0.30 C, 1.09 Cr and 0.24 Mo., in the quenched and tempered condition, of approximately 39% for the plated substrate and

U

Corresponding author. Tel.: q58-2-6628-927; fax: q58-2-7539017. E-mail address: [email protected] ŽE.S. Puchi Cabrera..

a reduction of 20% when the substrate was previously shot peened before plating. Also, Zhang et al. w2x, working on the same substrate, reported that plating it with an EN deposit reduced the fatigue limit of the material in comparison with the unplated substrate, although such a decrease was observed to be less marked if the substrate was previously shot peened. In both investigations, the low fatigue strength of the coating was found to be responsible for the decrease in the fatigue limit of the plated steel and that for this type of composite material, the fatigue properties depend primarily on the fatigue resistance of the coating itself. More recently, Garces ´ et al. w3x reported that plating a quenched and tempered AISI 4340 steel with

0257-8972r00r$ - see front matter 䊚 2000 Elsevier Science B.V. All rights reserved. PII: S 0 2 5 7 - 8 9 7 2 Ž 0 0 . 0 0 8 9 9 - 9

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

an EN deposit leads to a significant reduction of the fatigue life of the material that can reach up to 92%. The microscopic observation of the fracture surfaces of the samples conducted in this investigation indicated that the fatigue process was initiated at the surface of the deposit and, subsequently, transferred to the substrate, with the assistance of the metallic bonding established at the deposit᎐substrate interface. More important, this study have concluded that the EN deposit actually acted as a surface crack source or surface notch which decreased the fatigue life of the coated material by reducing the crack nucleation stage. Thus, the present investigation has been conducted with two different purposes. Firstly, to corroborate the hypothesis put forward above, by comparing the fatigue curves that are obtained in a quenched and tempered SAE 4140 steel, both in a notched and an unnotched condition, uncoated and coated as-deposited with a commercial EN plating of 20 ␮m thickness and a P content ranging approximately between 12 and 13 wt.% deposited industrially. Secondly, to test a method for obtaining an approximate estimate of the fracture toughness of the substrate material as well as the material parameters involved in the Paris relationship for the description of the fatigue crack growth rate as a function of the stress intensity factor.

2. Experimental techniques The present investigation has been carried out with samples of a SAE 4140 steel of the following composition Žwt.%.: 0.39 C, 0.75 Mn, 0.24 Si, 0.24 Cu, 0.95 Cr, 0.17 Mo and 0.15 Ni. This material is widely used in the automotive industry, for the manufacture of connecting rods, crankshafts, knuckles, rear axle and trailer axle shafts. Also, in the aircraft industry it is employed for making shapes and tubing, and in the oil industry for the production of bits, core drills, reamer bodies, drill collars, tool joints, piston rods and pump parts. The material was provided as bars of approximately 14 mm diameter and 6 m length. Such bars were cut to pieces of approximately 120 mm length for machining tensile specimens and of 90 mm length for machining the fatigue samples. Both type of specimens had a gage diameter of 6.35 mm, gage length of 12.7 mm, fillet radius of 25.4 mm and shoulder diameter of 12.7 mm, according to the ASTM standard E-606. The alloy was already provided in the quenched and tempered condition. The specimens were subsequently ground with successive SiC papers grit 80᎐1200 and polished mechanically to a mirror-like finish. Sixty fatigue samples were notched by spark machining employing a copper electrode, applying a potential of 45᎐50 V for 2 min. The stress concentrators thus produced were of a hemispherical shape with a radius of 0.5 mm. In order to

573

maintain this geometry, the electrode was remachined every three notches. The samples to be plated were degreased in a 5% HCl solution at 348᎐353 K for 7 min, rinsed in distilled water, rinsed in a sodium bicarbonate Ž100 grl. solution and rinsed again in water. The EN deposition was conducted industrially at Tecnologıas ´ Aplicadas C. A. ŽSan Diego de los Altos, Venezuela., employing a bath composed of 30 grl of nickel sulfate, 30 grl of sodium hypophosphite, 35 grl of malic acid, 1.5 ppm of lead sulfate, 10 grl of succinic acid and a stabilizer. During deposition the pH was maintained at approximately 4.6᎐4.8, at a mean temperature of approximately 363 K. The deposition rate was of approximately 12 ␮mrh. The deposit applied had a phosphorous content of the order of 12᎐13 wt.% and a thickness of approximately 20 ␮m, which was corroborated by means of the ball cratering technique ŽCalotest, CSEM. and image analysis ŽLECO 500.. The chemical analysis of the plating in the as-deposited condition was determined by means of SEM techniques ŽHitachi S-2400. with EDS facilities. The observations were conducted at a constant potential of 20 kV. Tensile tests were carried out on a computer-controlled servohydraulic machine ŽInstron 8502, USA. at a cross head speed of 3 mmrmin. At least three samples were employed for characterizing the monotonic mechanical properties of both the coated and uncoated substrate. All the fatigue tests were carried out under rotating bending conditions ŽFatigue Dynamics, RBF-200, USA. at a frequency of 50 Hz and alternating stresses of 474, 510, 545 and 580 MPa, which corresponded to approximately 50, 54, 58 and 62% of the yield stress of the unplated substrate. A total of 24 samples were employed for evaluating the fatigue properties of the material under the four different conditions investigated. The number of fatigue samples employed to determine the fatigue life curves fulfills the ASTM standard 739 Ž12᎐24 samples. for reliability data required in S᎐N testing. Thus, the testing procedure employed in the present investigation allowed a replication of more than 80%. The meaningful comparison of the fatigue life of the materials under different conditions was possible by machining and polishing all the specimens in order to have similar mirror-like polished surfaces before testing. SEM techniques were employed for the examination of the fracture surfaces of the samples, especially regarding three important aspects of the present investigation: Ža. the role of the EN deposit in the nucleation of cracks in the unnotched coated specimens; Žb. verification of the stress concentrator dimensions; and Žc. estimation of the critical crack size, at each stress level, at which the final fracture of the fatigue samples occurred.

574

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

3. Results and discussion 3.1. Characteristics of the deposit Fig. 1 shows a detailed view of the substrate ᎐deposit interface between the EN coating and the base steel of one of the samples after fracture. In this picture, some secondary cracks can also be observed along the interface, which possibly indicates a poor adhesion between both materials. Also, it is clearly observed that the fracture features generated during the fatigue process are shared between the deposit and the substrate. Some fatigue striations are noticed within the plating. The thickness of the coating determined from the photomicrograph, of approximately 20 ␮m, agrees with the measurements conducted by means of the ball cratering technique. Fig. 2 illustrates one of the EDS analyses conducted on the deposit, which allowed to determine a P content in the range of approximately 12᎐13 wt.%. Such results, according to Parker and Shah w4x, would indicate the existence of a compressive residual stress pattern within the coating. The results reported by Wu et al. w1x and Zhang et al. w2x would also corroborate this assumption. 3.2. Mechanical properties The influence of the EN deposit employed in the present study on the monotonic mechanical properties of the composite coating᎐substrate material was evaluated by conducting a number of tensile tests with samples both in the uncoated and coated conditions. For the substrate material, the yield stress was found to be approximately 817 " 6 MPa, whereas the ultimate tensile strength ŽUTS. was found to be approximately

Fig. 1. SEM view of the substrate ŽS. ᎐deposit ŽD. interface between the EN coating and the base steel after fracture. Secondary cracks ŽSC. can also be observed along the interface. It is clearly noticeable that the fracture features generated during the fatigue process are shared between the deposit and the substrate. Some fatigue striations ŽFS. are seen within the plating.

941 " 2 MPa. In the as-deposited condition the material had a yield stress of 819 " 2 MPa and a UTS of 939 " 2 MPa, which indicates that the deposit plated onto the substrate does not give rise to any change either in yield stress or in the UTS. Such a result is not surprising given the small thickness of the coating in relation to the diameter of the sample. As reported by Garces ´ et al. w4x, during testing of the coated samples, the deposits were observed to detach severely from the substrate, indicating a poor bonding at the substrate ᎐deposit interface and the brittle nature of the coating. As far as fatigue testing is concerned, the evaluation of the monotonic mechanical properties of the material allowed to determine a stress amplitude range of

Fig. 2. Typical EDS spectrum for the EN deposits involved in the present work.

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582 Table 1 Mean number of cycles to failure Ž Nf . vs. stress amplitude Ž S . for the substrate samples

Table 3 Mean number of cycles to failure Ž Nf . vs. stress amplitude Ž S . for the coated specimens

Stress ŽMPa.

Mean S.D.

575

Stress ŽMPa.

474

510

545

580

474

510

545

580

1 112 300 737 100 596 100 544 800 778 200 4 697 500 1 411 000 1 480 948

678 600 628 600 412 700 920 000 184 500 200 500 504 150 265 077

98 700 219 000 251 200 154 300 115 400 171 100 168 280 53 680

80 200 109 300 87 000 44 600 107 000 76 500 84 100 21 607

46 500 103 900 74 300 45 900 215 900 123 100 101 600 58 332

92 200 32 800 61 400 61 200 35 800 90 800 62 367 23 384

84 600 64 200 43 400 40 300 68 800 27 900 54 867 19 305

44 100 36 100 37 700 55 600 21 400 32 100 37 833 10 495

474᎐580 MPa to conduct the fatigue tests of both coated and uncoated materials. Such stress interval corresponded to a fraction of approximately 0.50᎐0.62 of the yield strength. Tables 1᎐4 present the data of number of cycles prior to fracture Ž Nf . in terms of the alternating stress applied to the material Ž S . for all the conditions investigated. Fig. 3 illustrates the results obtained from which it can be seen that at least six tests were conducted at each alternating stress. It has already been mentioned that this amount of samples allowed the fulfillment of the reliability conditions prescribed in the ASTM standard E-739. Several important aspects must be discussed in relation to Fig. 3. Firstly, the linear relationship between the alternating stress and the number of cycles to failure in a double logarithmic scale for all the conditions analyzed indicates the validity of the simple parametric expression of the type earlier proposed by Basquin w5x for the description of this kind of data:

Mean S.D.

where A and m are constants that depend on both material properties and testing conditions; A represents the fatigue strength coefficient of the material and m the fatigue exponent. Table 5 summarizes the values of the parameters A and m for the four sets of

data represented in Fig. 3. Such parameters would be of upmost importance for the design of structural components and parts that could fail by high cycle fatigue under service. This is particularly relevant for those parts made of high strength steels that are coated with EN deposits either as a mean of restoring their dimensions after severe wear in service or to improve both their corrosion or abrasive wear resistance before going into service. Also, it can be observed from Fig. 3 that plating an EN deposit of these characteristics onto the 4140 substrate significantly decreases the fatigue life of the material in relation to the uncoated steel, in spite of the fact that the deposit is under compressive residual stresses. At an alternating stress of 580 MPa the reduction in fatigue life reaches 68% whereas at 474 MPa it reaches 88%. These results, are consistent with those obtained by Wu and co-workers w1x and also by Zhang et al. w2x regarding the decrease in the fatigue limit of the 30CrMo steel, up to 52%, when it is plated with an EN deposit. Similarly, the present results agree with those reported by Garces ´ et al. w3x, for an AISI 4340 steel coated with an EN deposit, in the as-deposited condition, for which the decrease in fatigue life varied between 49 and 78% when the fatigue tests were conducted at alternating stresses in the range of 663᎐590 MPa.

Table 2 Mean number of cycles to failure Ž Nf . vs. stress amplitude Ž S . for the uncoated notched specimens

Table 4 Mean number of cycles to failure Ž Nf . vs. stress amplitude Ž S . for the notched coated specimens

S s ANfm

Ž MPa .

Ž1.

Stress ŽMPa.

Mean S.D.

Stress ŽMPa.

474

510

545

580

474

510

545

580

92 100 50 300 61 900 71 700 54 100 47 100 62 287 15 367

44 600 57 200 33 800 29 000 38 700 58 800 43 683 11 182

67 900 43 000 28 300 31 500 33 500 18 800 37 167 15 492

28 400 19 500 40 700 30 500 15 600 17 500 25 367 8778

48 200 41 700 37 900 72 800 86 100 33 100 53 300 19 416

32 000 21 800 24 700 54 300 21 600 53 100 34 583 13 952

19 800 26 100 45 200 23 200 14 800 21 700 25 133 9612

20 800 16 800 19 300 15 900 39 800 23 700 22 717 8058

Mean S.D.

576

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

Fig. 3. Mean number of cycles prior to fracture Ž Nf . as function of the alternating stress applied to the material Ž S . for the uncoated, uncoated notched, coated as-deposited unnotched and coated notched specimens.

As expected, the presence of a hemispherical notch in the uncoated samples also gives rise to a significant reduction in the fatigue life of the substrate material since the time required for the nucleation of the crack is virtually suppressed. The decrease in fatigue life varies between 69 and 94% in the alternating stress range employed in this investigation. It is interesting to observe that at 580 MPa the reduction in fatigue life induced by the stress concentrator is comparable to that found for the coated samples in the as-deposited condition. At low alternating stresses Ž474 MPa., the notch gives rise to a reduction in the fatigue life slightly greater Ž6%. than that produced by the EN deposit. These observations lead to the conclusion that the EN deposit acts as a surface notch or stress concentrator either because its lower mechanical properties enhance the nucleation of cracks that are subsequently transferred to the substrate or due to the presence of cracks within the deposit that were nucleated during its synthesis, which reach the critical size for propagation after few cycles of loading. If the notched samples are also coated with the EN deposit, the fatigue life of the substrate is decreased

Table 5 Parameters involved in the Basquin relationship for the conditions tested Condition

A ŽMPa.

m

Substrate Coated as-deposited Uncoated notched Coated notched

1119.5 1336.2 1850.9 1560.2

0.060 0.085 0.120 0.105

further, particularly at elevated alternating stresses where such a reduction can reach up to 78%. At low alternating stresses the decrease in fatigue life reaches approximately 95%, which is marginally superior than that induced by the stress concentrator alone, but within the experimental scatter of the results. Under low and elevated alternating stresses, the notch is the site initiation of the dominant crack that gives rise to the final fracture of the sample and therefore, as discussed later, the fracture surfaces of the notched specimens, both uncoated and coated, display a single crack site initiation and a flat surface. The above results could be interpreted more clearly in terms of the estimated number of cycles required for the nucleation of a fatigue crack of 0.5 mm, which at each alternating stress would be given by the difference between the number of cycles to fracture of either the uncoated or coated substrate and that corresponding to the notched uncoated specimens, which represents the number of cycles required for the propagation of such a crack. The relevant data, based on the mean of the number of cycles, is reported in Table 6 and 7 and shown graphically in Fig. 4a,b for both the uncoated and coated substrate, respectively. As it can be observed from Fig. 4a, the estimated number of cycles for the nucleation of a crack of this size in the uncoated substrate is very similar to that required for fracture, which means that for the uncoated material, most of the fatigue life is spent in the nucleation of the crack rather than propagating it. According to Table 6, at an alternating stress of 580 MPa, approximately 70% of the fatigue life is consumed in the nucleation of the crack, whereas at 474 MPa it takes approximately 96%. On the contrary, for the coated material, as shown in Fig. 4b and Table 7, most of the fatigue life is consumed in the propagation of the fatigue crack. At an alternating stress of 580 MPa, only 40% of the fatigue life is expended in the nucleation of the crack, whereas at 474 MPa such a process requires approximately 47.5%. Thus, these results show clearly that by plating the substrate material with the EN deposit, the nucleation of fatigue cracks is accelerated significantly and, therefore, that the coating acts effectively as a notch. Table 6 Mean crack length at fracture as a function of the alternating stress and mean number of cycles to fracture for the uncoated notched samples Stress ŽMPa.

Mean number of cycles of propagation Ž Np .

Mean critical crack length Žmm.

KIc ŽMPa m1r 2 .

474 510 545 580

62 867 43 683 37 167 25 367

3.69 3.49 3.32 3.15

62.0 61.9 62.0 61.9

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

577

Table 7 Mean number of cycles to fracture for the coated notched samples Stress, ŽMPa.

Mean of total number of cycles

Mean number of cycles of propagation

Mean number of cycles of nucleation

474 510 545 580

101600 62367 54867 37833

53300 34583 25133 22717

48300 27784 29734 15116

3.3. Fracture toughness of the substrate

⌬ Ks f Ž a. ⌬ S'␲ a

Ž2.

which, at fracture can be simply expressed as: The fracture toughness of the substrate material can be estimated from the classical expression that relates the stress intensity factor, at the crack tip, ⌬ K, with stress applied, ⌬ S, and the crack length, a:

K Ic s f Ž a c . Smax ␲ a c

'

In the above equations, f Ž a. represents a geometrical factor that depends on the applied loads and geometry of the body and crack. However, due to the complexities of the problems, exact solutions for edge cracks in rods under bending are not available w6x. For example, according to Toribio and co-workers w7x, for a hemispherical notch in a solid bar subjected to bending: f Ž a. s 0.821y 0.486

Fig. 4. Ža. Estimated number of cycles for the nucleation and propagation of a crack of 0.5 mm in the uncoated substrate. Žb. Estimated number of cycles for the nucleation and propagation of a crack of 0.5 mm in the coated substrate.

Ž3.

ž da / q 2.003 ž da /

2

Ž4.

where d represents the diameter of the bar. Si w6x has also proposed a solution for determining f Ž a. by combining selected solutions for curved and straight fronted cracks previously published. In this work, f Ž a. is given in a table as a function of the ratio ard. Particularly, for 0 F ardF 0.6, it is reported that 0.74F f Ž a. F 1.5. Thus, by measuring experimentally the critical crack length from the fracture surfaces of the samples it is possible to estimate the K Ic of the material. The SEM observations conducted on the fracture surfaces of the uncoated and coated notched samples, tested at different alternating stresses, allowed to measure the mean critical crack length as a function of the stress applied and the mean number of cycles to fracture, as shown in Table 6. Such data, together with Eqs. Ž3. and Ž4., yielded an estimation of the fracture toughness of the substrate of approximately 62 MPa m1r2 . Such a result agrees reasonably well with the value reported by Le May and Shaw w8x for this material of 66 MPa m1r2 , after tempering at 673 K. In order to determine the fracture toughness of the material by means of any of the standard test methods developed for this purpose w9x, it is necessary to specify the specimen dimensions on the basis of an approximate value for K Ic , in order to fulfill the condition that such dimensions must be sufficiently large in comparison with the plastic zone dimensions at the crack tip. Therefore, it is concluded that the present experimental approach could be em-

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

578

ployed for determining satisfactorily a reasonable estimate of K Ic . 3.4. Crack propagation rate The early work conducted by Paris and Erdogan w10x showed that the rate of crack propagation is related to the stress intensity factor by means of a simple parametric equation of the form: da p sCŽ⌬ K . dN

Ž5.

where C and p represent material properties. Forman and co-workers w11x, in an attempt to improve the correlation between the crack growth rate and ⌬ K, proposed an alternative relationship of the form: r

da BŽ⌬ K . s dN Ž 1 y R1 . K Ic y ⌬ K

Ž6.

where, again, B and r represent material parameters, R1 the load ratio and K Ic the fracture toughness of the material. The test method for conducting fatigue crack growth measurements is fully described in the ASTM E 647-88a standard and basically it consists in growing the crack by cyclic loading and monitoring K min , K max and the crack length throughout the test. The test fixtures and specimens design are similar to those employed for fracture toughness testing. However, a rough estimate of the materials constants involved in Eqs. Ž5. and Ž6. can be obtained from the fatigue experiments conducted with the uncoated notched and unnotched samples. As it has been shown before, due to the stress concentrator in these specimens, the number of cycles to fracture is approximately equal to the number of cycles for the propagation of the crack. Thus, by considering that under rotating bending conditions the propagation of the crack occurs due to the action of the maximum tensile stress at the crack tip, and therefore ⌬ Kf f Ž a. Smax Ž ␲a.1r2 , Eq. Ž5. can be combined with Eq. Ž4. and integrated directly to give: Np s

1 C

ac

da

0

p Ž f Ž a. '␲ a .

Ha

Syp

Ž7.

where Np represents the number of cycles during the propagation of the crack, a0 the initial crack length Ž0.5 mm., ac the critical crack length at fracture and S the maximum alternating stress applied to the specimen during testing. The data presented in Table 6 can be employed to determine the value of the constants C and p by means of non-linear regression analysis. Thus, by defining the error sum of squares as:

Fig. 5. Ža. Typical fracture surface of an uncoated notched sample tested at 474 MPa. The radial lines ŽRL. indicate that the crack started to propagate from the notch root ŽN.. Žb. Detailed view of the previous picture at the notch. Some of the radial lines actually correspond to secondary cracks ŽSC. that were also nucleated at this site. N

␸s

Ý is1

½

1 Np i y C

2

ac i

Ha

0

da Ž f Ž a. '␲ a .

p

Syp i

5

Ž8.

and solving the system of equations derived from the condition that: ⭸␸ s0 ⭸C

and

⭸␸ s0 ⭸p

Ž9.

In Eq. Ž8. N represents the number of experimental data available. Thus, it has been determined that for the substrate material C s 3.15= 10y1 1 mmrcycle and ps 4.22. In the above equations, ac is input in mm and S in MPa. If the solution for f Ž a. proposed by Si w6x is employed instead of that of Toribio and co-workers w7x, it is obtained that C s 1.56= 10y1 0 mmrcycle and p s 4.24. Such values are very close to those reported by Lemaitre and Chaboche w12x for similar materials. 3.5. Fracture surfaces of the samples Several specimens tested at different alternating stresses were examined after failure by SEM in order

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

Fig. 6. Ža. General fracture surface of a coated sample tested at 474 MPa showing the origin ŽO. of the crack. Žb. Closer analysis of the deposit illustrating the presence of fatigue striations ŽFS. within it.

to verify the size of the notch machined, to measure the crack length at fracture and also to study more closely the sites of crack initiation, particularly in the coated specimens. For example, Fig. 5a illustrates a photomicrograph of a typical fracture surface of an uncoated notched sample tested at 474 MPa. As expected, the radial lines indicate that the crack started to propagate from the notch root. The photomicrograph in Fig. 5b, which represents a more detailed view of the previous picture at the notch, indicates that some of the radial lines actually correspond to secondary cracks that were also nucleated at this site. The photomicrograph of Fig. 6a illustrates the general fracture surface of a coated sample tested at 474 MPa in which the origin of the crack is also well defined. A close analysis of the deposit, Fig. 6b, allowed to determine fatigue striations within it, which confirm our view in the sense that the EN deposit is able to develop fatigue cracks that subsequently are transferred to the substrate. Fig. 7a depicts the general fracture surface of a notched and coated sample, tested at 474 MPa. As for the uncoated specimens, it is observed that the fracture process occurs as a consequence of the propagation of the crack nucleated at the notch, giving rise to a flat fracture surface. Fig. 7b illustrates a detailed view of the notch root from which

579

Fig. 7. Ža. General fracture surface of a sample coated and notched, tested at 474 MPa showing that the fracture process occurs as a consequence of the propagation of the crack nucleated at the notch ŽN., giving rise to a flat fracture surface. Žb. Detailed view of the notch root showing the nucleation of several secondary cracks ŽSC..

several secondary cracks were also observed to originate. On the other hand, Fig. 8 illustrates a detailed view of Fig. 1, particularly of the fatigue striations that were observed within the EN deposit. These observations agree with the results previously reported by Zhang et al. w2x who were able to find fatigue markings within the EN deposits plated onto similar high strength steels as substrates.

Fig. 8. Detailed view of Fig. 1, showing fatigue striations ŽFS. within the EN deposit ŽD..

580

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

Finally, Fig. 9a shows the general fracture surface typically observed in the notched and coated samples tested at 580 MPa. At elevated alternating stresses it can be seen that the morphology of the fracture surface is quite similar to that observed at low stresses. It

can be clearly noticed that the main crack was nucleated at the notch root. Fig. 9b illustrates a closer look of the notch root, indicating a number of fracture steps which characterize the transcrystalline propagation of the fatigue crack.

Fig. 9. Ža. General fracture surface of a notched and coated sample tested at 580 MPa. The main crack was nucleated at the notch ŽN. root. Žb. Closer look of the notch root indicating a number of fracture steps ŽFST. which characterize the transcrystalline propagation of the fatigue crack. At the inner surface of the notch, a number of nodules ŽND. are observed to be present. Žc. Magnified view of zone A in Ža., illustrating the partial detachment of the deposit from the substrate through secondary cracking ŽSC.. Fatigue striations ŽFS. within the coating and fatigue marks ŽFM. on the substrate are also clearly visible. Žd. Magnified view of zone B in Ža., showing fatigue marks on the substrate. Most of the substrate ᎐deposit interface is observed to remain relatively free of secondary cracks.

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

At the inner surface of the notch, a number of nodules formed during the synthesis of the deposit are observed to be present, which give rise to an ‘orange-peeling’ type of surface finish all over the surface of the coated specimens. Hardness indentations conducted on such nodules allowed the determination of their solid nature. The zone on the left of the notch surface, designated as A, has been magnified in Fig. 9c. Here, the partial detachment of the deposit from the substrate through secondary cracking can be clearly observed, as well as fatigue striations within the coating and fatigue marks on the substrate. The zone on the top of the notch surface, designated as B, has also been magnified in Fig. 9d where fatigue marks are clearly visible on the substrate. In this area, most of the substrate ᎐deposit interface is observed to remain relatively free of secondary cracks. Thus, the present results corroborate those previously discussed regarding the fatigue tests conducted on coated and uncoated samples in the sense that EN deposits are bound to undergo fatigue failure before the substrate and to allow the transference of the fatigue cracks to it, giving rise to a reduction in its fatigue strength, particularly if the substrate has better mechanical properties than the deposit, as in the present case. Such a detrimental effect is almost comparable to the presence of a notch or stress concentrator from which the fatigue cracks nucleate and propagate. However, in the case of the coated materials the situation could be even worse than in the uncoated notched samples since it has been observed that at elevated stresses the presence of the deposit could accelerate the nucleation of the cracks that lead to the final fracture of the specimen.

4. Conclusions A number of fatigue experiments conducted on samples of a quenched and tempered SAE 4140 steel showed that plating this material with an EN deposit led to a significant reduction in the fatigue life. It has been determined that if the coating is in the as-deposited condition, at a stress amplitude of 580 MPa, such a reduction can reach up to approximately 68%, whereas at 474 MPa it could reach up to approximately 88%. The experiments conducted on the uncoated notched samples revealed that the presence of a stress concentrator of 0.5 mm in length, gave rise to a reduction in fatigue life that varied between approximately 69 and 94% depending on the alternating stress applied to the material. This decrease in fatigue properties is comparable to that reported for the coated specimens, which suggest that the EN deposit effectively acts as a surface fatigue crack source, causing a significant decrease in the fatigue properties of the substrate mate-

581

rial. If the notched samples are further coated, then at elevated alternating stresses the reduction in fatigue life is increased up to approximately 78%, whereas at low stresses it remains at a similar value to that observed for the notched and uncoated specimens. The analysis of the fracture surfaces of the coated and uncoated notched samples tested, revealed that at low alternating stresses, the dominant crack responsible for the fracture process is nucleated at the root of the notch, regardless of the presence of the deposit. Such study also revealed clear evidence of fatigue striations within the deposit and the continuity of certain fracture features between coating and substrate. Therefore, it has been concluded that the decrease in fatigue life in the as-deposited samples occurs as a result of the passage of fatigue cracks from the coating to the substrate. Extensive secondary cracking along the coating᎐substrate interface has been revealed after fatigue testing, which indicates that the adhesion of the EN deposit is somewhat poor and therefore the interface is not able to sustain the stresses applied to the material. The experiments conducted with the notched and uncoated specimens, together with the experimental measurement of the critical crack length at different alternating stresses, allowed to conduct a rough estimation of the fracture toughness of the substrate material which agrees satisfactorily with the values reported in the literature for this important property. Also, based on the information provided by the results obtained from the notched samples, a non-linear regression method has been presented by means of which it has been possible to estimate satisfactorily the parameters that enter in the Paris relationship for the description of the crack growth rate of the substrate, in terms of the stress intensity factor at the crack tip. The two methods advanced here, which can be easily implemented, can provide both useful and reliable information about the fracture properties of the material being tested before conducting the more elaborated procedures already standardized.

Acknowledgements This investigation has been conducted with the financial support of the Venezuelan National Council for Scientific and Technological Research ŽCONICIT. through the project LAB-97000644 and the Scientific and Humanistic Development Council of the Central University of Venezuela ŽCDCH-UCV. through the project 08-17-4595-2000. The authors would like to acknowledge the assistance of Mr E. Batoni and B. Lozada in the conduction of the experimental work. J.A. Berrıos ´ is deeply grateful to the School of Mechanical Engineering, Faculty of Engineering and Architecture of the University of El Salvador.

582

A. Pertuz et al. r Surface and Coatings Technology 133᎐134 (2000) 572᎐582

References w1x Y.Y. Wu, Y.Z. Zhang, M. Yao, Plating Surf. Finish. Ž1995. 83᎐85. w2x Y.Z. Zhang, Y.Y. Wu, M. Yao, J. Mater. Sci. Lett. 15 Ž1996. 1364᎐1366. w3x Y. Garces, J. Berrıos, ´ H. Sanchez, ´ ´ A. Pertuz, J. Chitty, H. Hintermann, E.S. Puchi, Thin Solid Films 355r356 Ž1999. 487᎐493. w4x K. Parker, H. Shah, Plating 58 Ž1971. 230. w5x O.H. Basquin, Proc. ASTM 10 Ž2. Ž1910. 625. w6x E. Si, Eng. Fract. Mech. 37 Ž4. Ž1990. 805. w7x J. Toribio, V. Sanchez, M. Astiz, J. Campos, J. Eng. Fract. ´ Mech. 39 Ž1991. 357.

w8x I. Le May, J.D. Shaw, in: J. Backlund, A.F. Blom, C.J. Beevers ŽEds.., Fatigue Thresholds: Fundamentals and Engineering Applications, v噛ol. II, Engineering Materials Advisory Services Ltd, The Chameleon Press Ltd, London, UK, 1982, p. 603. w9x E 399-83, Standard test method for plane᎐strain fracture toughness of metallic materials, American Society for Testing and Materials, Philadelphia, 1983. w10x P.C. Paris, F. Erdogan, J. Basic Eng. Trans. ASME Ser. D 85 Ž4. Ž1963. 528. w11x R.G. Forman, V.E. Kearny, R.M. Engle, J. Basic Eng. Trans. ASME 89 Ž1967. 459. w12x J. Leimatre, J.L. Chaboche, Mechanics of Solid Materials, Cambridge University Press, New York, 1994, p. 484.