Effect of Shot Peening and Nitriding on Toughness and Abrasive Wear Resistance of Powder Metallurgic Steels Highly Alloyed with Vanadium

Abstract

Böhler K390 steel is used for cold work tools, with 9% of V, made by using powder metallurgy. In this work, it has been studied the effect of shot peening and nitriding surface treatments on wear resistance and impact toughness of this type of steel. For this purpose, previous changes in several thermal processing factors related to quenching and tempering were carried out. The results allow for an increase in the hardness, impact toughness, and abrasive wear resistance of these steels. An austenitizing treatment at 1100 °C with air cooling and 3 tempering processes at 550 °C is suggested. These conditions foster a lower weight percentage of retained austenite, up to 3%, a higher carbide percentage, up to 15–16% in weight, and a greater impact toughness with no notch, of above 40 J/cm². If this treatment is combined with further ion nitriding, the maximum level of abrasive wear resistance is reached. The only carbide type present in the microstructure is the MC type. Most of the V, Cr, and Mo contents are present in said carbides. The Co and the W tend to remain in solid solution in the matrix constituent. Both the shot peening treatment as well as ion nitriding offer a considerable increase in hardness, with values of up to 1500–1600 HV. Nevertheless, it has been confirmed that shot peening does not offer any abrasive wear resistance improvement. Such resistance may only be considerably improved by the application of an ionic nitriding treatment. The thickness of the nitrided layer fluctuates between 150 and 175 µm. The carbides are affected by nitriding, reaching levels that are higher than the atomic 10%, at an intermediate depth of the nitrided layer. These values are higher in the matrix constituent, as they are even higher than the atomic 20% in N.

1. Introduction

The cold work tool steels are highly used in the metallurgical industry as cutting tools, molds, punches, dies, etc. [1,2]. This type of steel is required to have highly strong properties such as a high level of toughness and wear resistance [3]. These steels are used in quenched and tempered conditions. Manufacturing of this type of steel using powder metallurgy (PM) gives the possibility to use chemical compositions that could not be used for steel manufacturing if conventional casting processes were implemented. This manufacturing technique gives the possibility to avoid micro-segregation of alloy elements. Thus, the Ms temperature is higher and homogeneous, reducing the existence of retained austenite. In traditional casting processes, elements such as C and Cr, which are common in tool steels, tend to segregate into the residual liquid phase, which favors a decrease of the Ms temperature in these zones and, therefore, the presence of retained austenite [4]. At the same time, this technology allows the manufacture of parts with dimensions that have narrow tolerance margins [5]. On the other hand, steels manufactured by powder metallurgy have a more homogeneous distribution of carbides, thus fostering wear resistance in comparison with traditional manufacturing techniques, such as forging [6,7].

These steels have a high content of C and the most frequent alloy elements are Cr, Mo, V, and W [8]. Hardened microstructure of this type of steel is mainly made up of martensite, with variable amounts of retained austenite and carbides [9]. Microstructure may include primary carbides of the M₆C type, related to W, carbides of the MC type related to V, carbides of the M₇C₃ type related to Cr, and M₂C carbides related to Mo [10], which improve wear resistance features [11]. The presence of a high density of VC (vanadium carbide) increases adhesive wear resistance and Mo improves steel hardenability [12,13]. On the other side, the effect given by W may also be achieved by adding Mo. Considering that the Mo atomic weight is half of the W one, half of the Mo weight will be needed to keep the same atomic relation when replacing W with Mo [11].

Ion nitriding allows the improvement of the wear resistance of these steels. This treatment uses electric discharge in gases to produce plasma, thus fostering nitrogen interstitial insertion [14]. The existence of tempered martensite fosters N diffusion through the octahedral gaps of Fe-BCC [15]. N reacts with Fe and with other alloy elements creating two layers: a highly thin first layer made up of nitrides of the ε-Fe_2–3N type and of the γ’-Fe₄N type [16]. They produce an important distortion in the ferritic matrix [17]. Generally, this layer usually does not exceed 25 µm [16]. The nitrided layer has a second thicker layer of about 200 µm. The existence of alloy elements, such as Cr, Al, Ti, Mo, V, and W, supports the creation of nitrides in said layer [5]. During nitriding, nanoprecipitates of the CrN type, coherent with the surrounding matrix, may appear. They will then increase their own size, losing consistency with the matrix [14]. Nanoprecipitates of VN may also be formed with a thickening kinetics quite slower than the previous ones [18,19]. Carbides present in the microstructure might also absorb atoms of N [14].

Shot peening is one of the usual treatments used in industry to improve fatigue behavior, even in the case of tool steels and when a high level of wear resistance is required [20,21,22,23,24,25]. This surface treatment, apart from leading to a surface plastic deformation that produces a residual compression stress field (which results in an important surface hardening) may also foster the increase of wear resistance [26,27,28,29,30].

Most degrading mechanisms, such as wear, fatigue, and corrosion, begin on the surface [31]. At present, it is possible to improve the surface properties of tool steels by using ion nitriding or shot peening [1,31].

The Böhler K390 steel is a cold work tool steel, highly alloyed in V, manufactured by PM, and having high wear resistance, good ductility, and suitable toughness properties [6,31,32]. In this work, it has been studied the effect of surface nitriding and shot peening treatments over wear resistance and impact toughness of this type of steel. At the same time, the influence of several thermal processing factors on these properties was also analyzed.

2. Materials and Methods

By implementing an experiment design, it was intended to study the effect of different parameters of thermal processing of the Böhler K390 steel, manufactured by PM. The main variables analyzed were abrasive wear resistance, according to the ASTM G65 Standard [33], and impact toughness. It also analyzed the effect of submitting the material to surface nitriding and shot peening treatments, subject to the previous thermal treatment. The experimental methodology followed was that of a divided experiment design, with 5 factors and 8 experiments.

Table 1. Chemical composition of Böhler K390 steel (wt.%).

Composition	C	Cr	Mo	V	W	Co	Mn	Si
Böhler K390	2.5	4.2	3.8	9.0	1.0	2.0	0.4	0.5

shows the steel chemical composition,

Table 2. Factors and levels analyzed in the design of experiments.

Code	Factors	Level −1	Level +1
A	Austenitizing temperature	1100 °C (1 h)	1200 °C (1 h)
B	Cooling medium	air	oil
C	Tempering temperature	500 °C (2 h)	550 °C (2 h)
D	Amount of tempering	2	3
E	Surface treatment	Shot-peening	nitriding

shows the factors and levels analyzed and

Table 3. A matrix of experiments followed in the design of experiments.

Experiment	A	B	C	D	E	Confusion Pattern
1	−1	−1	−1	1	1	A+BD+CE B+AD C+AE D+AB E+AC BC+DE BE+CD
2	1	−1	−1	−1	−1
3	−1	1	−1	−1	1
4	1	1	−1	1	−1
5	−1	−1	1	1	−1
6	1	−1	1	−1	1
7	−1	1	1	−1	−1
8	1	1	1	1	1

shows the experiment matrix [34]. Four factors are related to the steel thermal processing (A to D codes). The other factor refers to surface treatments (E code).

The main purpose when following an experiment design is to deliberately modify the ordinary working conditions to produce changes in some of the responses being studied [35]. In this case, abrasive wear resistance and impact toughness. Such changes are made on specific factors, that are previously chosen. If 2 is the number of levels of each factor, and k is the number of factors, then 2^k will be the number of experiments to be carried out in a factorial experiment design [36]. In this case, k = 5, thus, if it were a factorial design, 32 experiments would be necessary. Fractional designs make it possible to reduce the number of experiments, at the expense of losing some information which, in industrial practice, tends not to be relevant. Fractional designs allow the number of experiments to be reduced, at the expense of losing some information which, in industrial practice, tends not to be relevant [36,37]. In this case, the number of experiments would be 8 = 2⁵⁻², where the number of “generators” is 2: Column D=AB and E=AC [36]. A factor effect is defined as the variation of the response function as a consequence of the variation of said factor between its low (−1) and high (+1) levels. This analysis agrees with the main effects. Interactions between two factors are defined as the variation between the average effect of one factor with the other one at its low level, −1, and the average effect of that same factor with the other factor at its high level, +1 [38]. Similarly, interactions between more than two factors will be defined [36]. The influence of main effects on the response function tends to be higher than the influence of 2-factor interactions and the latter tends to be higher than 3-factor interactions and so on. This justifies the division of an experiment design [36]. Nevertheless, it is important to analyze interactions of type II hidden behind a main effect. These type II interactions are included in the “confusion pattern” column of Table 3. A certain effect is defined as relevant if it is highly “rare” to take place by chance. The Pareto charts give the possibility to organize the effects analyzed based on their importance [36]. The experimental response is subject to random variation. This variation will follow an ordinary law, where its typical deviation reflects the experimental error. Effects are linear combinations of responses. Thus, according to the Central Limit Theorem, they will follow an ordinary law [34,35]. Each effect may be considered a random variable where the value obtained is an estimation of its average. If all the effects are insignificant, they will follow an ordinary zero average law, and thus, they will be aligned on a standard probability paper plot [38]. Significant effects follow an ordinary distribution with an average that is different from zero so they do not appear aligned with the insignificant ones. When one significant effect separates itself from the ordinary probability line towards the right side and on its upper part, the conclusion is that if this factor is placed at its +1 level, the response function will “significantly” increase with reference to its −1 level. Additionally, if a significant effect separates from the ordinary probability line towards the left side and on its upper part, the conclusion is that if this factor is at its −1 level, the response function with reference to its +1 level will “significantly” increase [34,35]. Statistical analysis has been carried out with the STATGRAPHICS Centurion XVI software (V16.2.04, Statgraphics Technologies, Inc., The Plains, VA, USA) [39,40,41,42].

Utilizing X-ray diffraction, the percentage of precipitated stages was established. For this purpose, the Seifert XRD 3000 T/T equipment (Baker Hughes, Celle, Germany) was used. Radiation was emitted through a Mo tube with a power of 40 kV and 40 mA. Measurements were made in the angular range of 2θ, between 8° and 37°, with an angular step of 0.03° and counting period per step of 22 s. The determination of the percentage of crystalline phases has been established by adjusting the diffragtogram using the Rietveld structural refinement method [43]. For this purpose, crystollographic information files of the ICSD (Inorganic Crystal Structure Database, FIZ Karlsruhe, Eggenstein-Leopoldshafen, Germany) were used. The software used was FullProf.2k, the 6.20 version [44,45]. The width increase observed in the peaks of the majority phases has been modeled using Stephens’s formulation, implemented in this analysis software [46].

Abrasive wear resistance was measured according to the G65 Standard. Rectangular section specimens measuring 55 × 30 mm and 5 mm thick were used. Experiment parameters were the following ones: rubber tyre diameter = 228 mm; rubber Shore A hardness = 60; silica sand (AF60/70); sand flow = 300–400 g/min; turning speed = 200 rpm; testing time = 10 min; applied load = 130 N.

Shot peening treatment was carried out on abrasive wear specimens detailed above, by using high-toughness ceramic projectiles diameter measuring 300 μm to 425 μm (ZIr Shot Y300) with an Almen intensity of 10 A. Treatment was carried out following the SAE J442 and J443 specifications. A nozzle with a diameter of 5 mm with a 230 mm distance between the nozzle and the working piece was used. Shot peening time corresponding to “5 s” for a minimum of 100% coverage. The impact angle was 90° and 100% coverage [47,48,49]. The coverage was analyzed with the NIS-Elements BR v5.20.02 software of the optical microscope. Figure 1 shows SEM images of the Zir Shot Y300 projectile used in the treatment.

In this research, the roughness was not considered because a high hardness 1000 HV projectile was used that breaks without leaving corners, small in size, between 300 µm and 425 µm, and does not increase the surface roughness significantly [48].

Table 4. Parameters used in the plasma nitriding process.

Gas mixture	70%N2 + 30%H2
Gas flux (cm3/min)	500
Temperature (°C)	540
Pressure (Pa)	400
Time (min)	120
Output voltage (V)	500

shows the main parameters of the process with which the plasma nitriding was carried out in experiments 1, 3, 6, and 8.

To determine Vickers hardness, the load implemented was 1000 N. To measure surface hardness, after surface treatments, the load implemented was 3 N. The Hardness value was estimated as the average value of 12 tracks.

The impact toughness (J/cm²) was determined by employing the Charpy test on non-notched specimens of square sections of 10 mm side and 55 mm length. The value for each experiment was estimated as the average value of 5 tests.

To disclose the microstructure, after the mechanical processes of grinding and polishing, Nital 4 was used as a chemical reagent. The optical microscope used was the NIKON Epiphot 200 (Nikon, Tokyo, Japan). The scanning electron microscope (SEM) used was JEOL JSM-5600, equipped with the characteristic dispersive X-ray microanalysis system (JEOL, Nieuw-Vennep, The Netherlands).

3. Results and Discussion

Figure 2 shows this steel microstructure in the annealed condition. Figure 2a shows high carbide density in a ferritic matrix. Material hardness in this condition is 280 HV. Figure 2b shows the points where the characteristic dispersive X-ray semi-quantitative microanalysis (EDX) was carried out.

Table 5. Carbide semi-quantitative analysis is specified in Figure 2. EDX semi-quantitative microanalysis (atomic%).

Spectrum	%C	%V	%Cr	%Fe	%Co	%Mo	%W
1	49.00	6.41	2.58	40.56	--	1.45	--
2	44.69	6.68	2.32	44.95	--	1.36	--
3	46.91	5.35	2.17	44.01	--	1.56	--
4	48.78	4.32	4.25	41.37	--	1.28	--
5	41.92	8.24	3.24	45.39	--	1.21	--
6	--	13.89	5.09	74.09	2.21	3.99	0.73
7	--	14.2	4.15	75.78	1.96	3.22	0.69

shows the results. Using the SEM-EDX technique it is not possible to determine with accuracy the atomic % of elements with low atomic numbers (as would be the carbon case). Therefore, the results shown in this table are semi-quantitative. These results are only intended to show the approximate distribution of alloying elements between the carbides and the matrix constituent in the annealed state. From spectra 1 to 5, it may be concluded that these carbides seem to be mixed V-rich carbides with MC stoichiometry [50]. Nevertheless, the matrix shows more V in solid solution, apart from Cr, Co, Mo, and W.

Table 6. Microstructural parameters, weight distributions of the precipitated phases.

No.	Rietveld Fitting	Phases	a (Å)	Error [a (Å)]	Lattice System	wt. %	Error [wt. %]
1	Rwp = 8.60	Fe(BCC)	2.87808	0.002	cubic	79.45	2.73
	Rexp = 5.22	Fe(FCC)	3.59529	0.005	cubic	4.67	0.92
	Chi2 = 2.71	VC carbide	4.17267	0.003	cubic	18.88	1.49
2	Rwp = 9.80	Fe(BCC)	2.87626	0.002	cubic	71.07	2.57
	Rexp = 5.07	Fe(FCC)	3.59724	0.003	cubic	16.35	1.41
	Chi2 = 3.74	VC carbide	4.16449	0.004	cubic	12.58	1.40
3	Rwp = 8.66	Fe(BCC)	2.87707	0.002	Cubic	77.23	2.46
	Rexp = 4.99	Fe(FCC)	3.59621	0.004	cubic	10.25	1.11
	Chi2 = 3.01	VC carbide	4.16832	0.003	cubic	12.52	1.22
4	Rwp = 9.77	Fe(BCC)	2.88037	0.002	cubic	71.21	2.58
	Rexp = 5.00	Fe(FCC)	3.60153	0.003	cubic	19.03	4.50
	Chi2 = 3.82	VC carbide	4.16906	0.002	cubic	9.76	1.21
5	Rwp = 8.18	Fe(BCC)	2.87492	0.002	cubic	80.55	2.71
	Rexp = 5.43	Fe(FCC)	3.59023	0.008	cubic	2.98	0.81
	Chi2 = 2.26	VC carbide	4.17417	0.003	cubic	16.47	1.41
6	Rwp = 8.10	Fe(BCC)	2.87498	0.0015	cubic	83.14	2.63
	Rexp = 5.22	Fe(FCC)	3.59025	0.008	cubic	3.40	0.80
	Chi2 = 2.41	VC carbide	4.16906	0.002	cubic	13.45	1.21
7	Rwp = 9.20	Fe(BCC)	2.87378	0.002	cubic	75.51	2.52
	Rexp = 5.17	Fe(FCC)	3.59684	0.004	cubic	10.67	1.20
	Chi2 = 3.16	VC carbide	4.16874	0.003	cubic	13.81	1.93
8	Rwp = 8.99	Fe(BCC)	2.87507	0.002	cubic	79.02	2.59
	Rexp = 5.09	Fe(FCC)	3.59597	0.004	cubic	9.46	1.12
	Chi2 = 3.12	VC carbide	4.16794	0.003	cubic	11.52	1.23

shows the weight percentages and the lattice parameters of the crystal phases observed by X-ray diffraction. It must be taken into account that, in all experiments, ferrite (tempered martensite) is the majority phase, and that the VC was the only carbide observed. Besides, it must be mentioned the existence of retained austenite in all the experiments.

Figure 3 shows the global settings obtained by the Rietveld structural refinement. The vertical segments indicate the angular positions of the different phases identified. Each phase is assigned a color.

Table 7. Average values and effects associated with austenite, VC, and ferrite contents.

Experiment	Austenite		VC		Ferrite		Effect
	(wt.%)	Effect	(wt.%)	Effect	(wt.%)	Effect
1	4.67	9.601	16.08	13.274	79.45	77.147	Average
2	16.35	4.917	12.58	−2.892	71.07	−2.07	A + BD
3	10.25	5.502	12.52	−2.742	77.23	−2.81	B + AD
4	19.03	−5.947	9.76	1.077	71.21	4.81	C
5	2.98	−1.132	16.47	0.367	80.55	0.82	D + AB
6	3.40	−5.312	13.45	0.237	83.14	5.12	AC
7	10.67	1.372	13.81	0.447	71.51	−1.77	BC
8	9.46	0.317	11.52	−0.002	79.02	−0.36	CD

shows the average values obtained in each experiment and the effects corresponding to the confusion patterns mentioned in Table 3. Considering that the X-ray diffraction analysis was carried out before the surface treatments (shot peening and nitriding), the E factor was not taken into account in the confusion patterns. The average shows the mean value obtained in the 8 experiments.

Figure 4 includes graphics showing the effects related to weight percentages of austenite, vanadium carbides, and ferrite (tempered martensite). Figure 4a,c,e show the Pareto charts, and Figure 4b,d,f, standardized effect representations on standard probability paper.

In the case of the austenite content, it seems that none of the analyzed factors has a significant effect. Nevertheless,

Table 8. Average percentages of austenite maintaining constant levels of factors A, B, and C.

Factor	Level	Austenite (wt.%)
A (austenitizing temperature)	−1 (1100 °C)	7.1
	+1 (1200 °C)	12.1
B (cooling medium)	−1 (air)	6.9
	+1 (oil)	12.4
C (tempering temperature)	−1 (500 °C)	12.6
	+1 (550 °C)	6.6

shows the average weight percentage of austenite based on destabilization temperature, cooling medium, and tempering temperature. It seems that high austenitizing temperatures (1200 °C) increase the retained austenite amount, as a consequence of higher carbon dissolution in austenite, which results in Ms temperature reduction. At the same time, air cooling allows carbide precipitation during cooling, and tempering temperatures of 550 °C foster a second austenite destabilization which gives rise to an additional transformation of austenite into martensite.

In the case of vanadium carbide content, it is found that the factors C (tempering temperature), A (austenitizing temperature), and B (quenching medium) have a significant effect. Within the analysis of the main effects of A and B, we have included the interactions between BD, “masked” in the effect of A, and AD, “masked” in the effect of B. To determine which of them has a significant “weight” on the VC content, the main effects and interactions must be carried out separately.

Table 9. Analysis of the A+BD and B+AD interactions, related to VC content.

(a). A + BD analysis
A	−1	+1	B(↓) × D(→)	−1	+1
−1	14.72		−1	13.01	16.27
+1		11.82	+1	13.16	10.64
(b). B + AD analysis
B	−1	+1	A(↓) × D(→)	−1	+1
−1	14.65		−1	13.16	16.27
+1		11.90	+1	13.01	10.64

shows the results. It is concluded that the significant effects are due to the C factor (tempering temperature) and the BD and AD interactions. At tempering temperatures of 550 °C, a second destabilization of the austenite occurs, which favors the precipitation of carbides of the MC type, associated with V. On the other hand, if during the thermal processing the austenitizing temperature, the cooling medium, and the number of tempering are simultaneously at their −1 (1100 °C), −1 (air cooling), and +1 (3 tempering) levels, the weight percentage increase of the VC takes place. It may be observed that experiments 1 and 5, which fulfill these requirements, are the ones with a high level of VC content, 16.08%, and 16.47% respectively. Lower austenitizing temperatures (1100 °C) benefit better austenite destabilization compared to 1200 °C and air cooling allows more carbide precipitation during cooling. 3 temperings give rise to a second austenite destabilization.

On the other hand, if we analyze the effects with reference to ferrite content (tempered martensite), significant effects may be observed in factor C (tempering temperature) and AC interaction (austenitizing and tempering temperatures). Tempering at 550 °C increases ferrite percentage and such percentage is increased if, together with the tempering at 550 °C, austenitizing is previously carried out at 1200 °C.

Table 10. Analysis of the AC interaction, related to ferrite (tempered martensite) content.

A(↓) × C(→)	−1	+1
−1	78.34	78.03
+1	71.14	81.08

shows a detailed analysis of the AC interaction. It is observed that, whenever austenitizing temperature is at 1100 °C, the ferrite-martensite percentage is high, without taking the tempering temperature into account. Nevertheless, if the austenitizing temperature were at 1200 °C, a lower level of ferrite-martensite formation would take place due to the high austenite percentage. However, this effect is balanced by a higher tempering temperature (550 °C), fostering this retained austenite destabilization. Therefore, it seems that tempering temperature is crucial. If, apart from the austenitizing temperature at 1200 °C, the tempering is carried out at 550 °C (experiments 6 and 8), the average ferrite percentage is 81.08%. However, if tempering is carried out at 500 °C (experiments 2 and 4), the average ferrite percentage decreases up to 71.14%, increasing that of retained austenite up to 17.7%.

Table 11. Ferrite (tempered martensite) lattice parameters and the effects associated with the pattern of confusion are indicated in the experimental matrix (Table 3).

Experiment	Fe(BCC) Lattice Parameter		Effect
	a (Å)	Effect (×10−3)
1	2.87808	2.87613	Average
2	2.87626	0.7	A+BD
3	2.87707	0.2	B+AD
4	2.88037	−3.2	C
5	2.87492	1.7	D+AB
6	2.87498	0.2	AC
7	2.87378	−1.2	BC
8	2.87507	−0.7	CD

shows the average lattice parameter of ferrite (tempered martensite) in each of the 8 experiments and the effects corresponding to the pattern of confusion indicated in the experiment matrix (Table 3). Figure 5 shows the graphical representation of the effects linked to this lattice parameter. Figure 5a shows the Pareto diagram and Figure 5b shows the standardized effect representations on standard probability paper. It can be seen that the only factor with a significant effect is the C factor (tempering temperature), so if we place this factor at its level −1 (500 °C) this lattice parameter increases, which would show a lower degree of tempering of the martensite.

Figure 6 shows the microstructure of experiments 4 and 5, showing the points where the semi-quantitative microanalysis was carried out by characteristic X-ray dispersion (EDX). Results are shown in

Table 12. Carbide semi-quantitative analysis is pointed out in Figure 6. Characteristic X-ray dispersion microanalysis (EDX). (atomic%).

Spectrum	%C	%V	%Cr	%Fe	%Co	%Mo	%W
1	60.14	15.89	2.62	18.39	--	2.96	--
2	59.81	14.56	2.6	21.2	--	1.83	--
3	52.33	22.77	3.31	17.74	--	3.85	--
4	44.5	5.75	3.04	44.99	--	1.72	--
5	--	3.00	5.13	87.18	2.21	1.4	1.08
6	50.44	21.37	3.22	21.23	--	3.74	--
7	54.4	26.01	3.16	12.46	--	3.97	--
8	--	5.49	5.03	84.42	2.55	1.68	0.83

. If these results are compared with those obtained in the annealed condition (Table 4), it seems that, after thermal processing, MC carbides have greater V enrichment and the V amount in the matrix solid solution has considerably been reduced.

Figure 7 shows the relative concentration of each element along the scanning line. V is the most abundant alloying element. However, it can be observed that this element is mainly concentrated in the carbides, as opposed to its solid solution state. The same behavior seems to be repeated for Cr and Mo. W and Co, however, seem to tend to be found in solid solutions in the matrix constituent.

Table 13. Results of Vickers hardness test. Hardness was determined before surface treatments (shot peening and nitriding) by applying a load of 100 kp. Surface hardness refers to the hardness measured on the surface of the parts after the surface treatments. In this case, a load of 0.3 kp was applied. Average values and effects are included. The margin of error is included with a confidence level of 95%.

Experiment	Hardness			Surface Hardness			Effect
	HV (100 Kp)	CL (95%)	Effect	HV (0.3 Kp)	CL (95%)	Effect
1	732	±16	745.6	1356	±9	1446.4	average
2	861	±7	47.2	1541	±12	96.7	A+BD+CE
3	778	±14	59.7	1511	±12	−105.7	B+AD
4	888	±8	−138.2	1582	±15	−101.2	C+AE
5	650	±13	−2.2	1670	±17	124.2	D+AB
6	620	±15	−72.2	1430	±11	−30.2	E+AC
7	728	±11	23.2	1055	±4	−202.7	BC+DE
8	708	±12	7.2	1428	±17	182.2	BE+CD

shows the hardness results. The first hardness column shows material hardness before surface treatments. The second column shows the hardness of the material surface after said surface treatments. Figure 8 shows the graphic of effects using Pareto charts and standard probability paper. Based on Figure 8a,b it may be concluded that factor C (tempering temperature) and the AC interaction (austenitizing and tempering temperatures) exert an important effect on hardness before the surface treatment has been carried out. This means that with tempering temperatures of 500 °C, hardness increases, and this could be due to a lower degree of tempering of the martensite. However, said hardness is “improved” if previous austenitizing is carried out at 1200 °C. Both austenitizing and tempering temperatures also result in a higher retained austenite content, that reaches an average value of 16.5% under these conditions (experiments 2 and 4). It is, thus, concluded that retained austenite with a low level of destabilization causes hardness increase.

Table 14. Impact toughness results by Charpy test on unnotched specimens. Average values and effects are included. Minimum and maximum values are shown in parentheses.

Experiment	Charpy Test			Effect
	J/cm2	(J/cm2)	Effect
1	33.91	(32.39–35.18)	24.74	Average
2	17.75	(16.73–18.76)	−11.90	A+BD
3	21.58	(17.13–25.44)	−11.97	B+AD
4	10.19	(9.65–10.81)	7.77	C
5	41.75	(38.10–46.69)	2.29	D+AB
6	29.51	(26.59–32.88)	−1.87	AC
7	25.54	(23.28–29.32)	−2.03	BC
8	17.71	(16.97–18.90)	−0.09	CD

shows the results obtained in the Charpy test with no notch. Figure 9 shows the graphic representation of effects using Pareto charts and standard probabilistic paper. The C factors (tempering temperature), A (austenitizing temperature), B (cooling mode) and the BD and AD interactions, which include the D factor (tempering number) have a significant effect.

Table 15. Analysis of A + BD and B + AD interactions, related to impact toughness.

(a). A + BD analysis
A	−1	+1	B(↓) × D(→)	−1	+1
−1	30.70		−1	23.63	37.83
+1		18.79	+1	15.88	13.95
(b). B + AD analysis
B	−1	+1	A(↓) × D(→)	−1	+1
−1	30.73		−1	23.56	37.83
+1		18.75	+1	23.63	13.95

shows the analysis of said interactions, and it demonstrates that the significant effect of the A factor also includes that of the BD interaction and the significant effect of the B factor also includes that of the AD interaction. The conclusion is that material impact toughness increases with austenitizing temperatures of 1100 °C, air cooling, together with 3 temperings, and tempering temperatures of 550 °C. Experiment 5 fulfils these requirements and it was the only one that exceeded the 40 J/cm². The austenitizing temperature of 1100 °C, air cooling and tempering at 550 °C are the levels at which less retained austenite is obtained. As 3 temperings are included, the conditions at which a higher percentage of precipitated VC carbides is obtained, are reproduced, i.e., a higher level of austenite destabilization.

Table 16. Abrasive wear resistance test results. Average values and effects are included.

Experiment	Loss Weight		Effect
	mg	Effect
1	18.3	27.12	average
2	49.3	4.3	A+BD+CE
3	17.2	−6.1	B+AD
4	30.7	−3.5	C+AE
5	35.5	−2.3	D+AB
6	17.7	−17.95	E+AC
7	28.9	3.7	BC+DE
8	19.4	6.4	BE+CD

shows the weight losses of the 8 experiments due to the abrasive wear test. Effects are included. Figure 10 shows the graphic representation of said effects. Figure 10a shows the graphic representation according to the Pareto chart and Figure 10b shows the graphic representation on standard probability paper. The only factor with a significant effect on abrasive wear resistance is surface treatment. Weight loss considerably increases if this factor is at its −1 level (shot peening). Thus, to increase wear resistance, the E factor should be at its +1 level (nitriding). The average weight loss of those experiments with shot peening (experiments 2, 4, 5 and 7) was 36.1 mg. However, the weight loss of those experiments submitted to nitriding (experiments 1, 3, 6 and 8) was 50% lower (18.2 mg). The conclusion is that the shot peening treatment has no positive effect on abrasive wear resistance.

Thicknesses of the nitrided layer were, in all cases, about 150 to 175 µm. Figure 11 shows an example of the distribution of nitrogen in the nitrided layer. It may be observed that N concentration is not homogeneous, as it considerably, decreases noticeably from the first 40 µm (approximately).

Figure 12 shows two representative pictures of the nitrided layer microstructure in an average area, located at about 80–90 µm from the sample surface.

Table 17. Semi-quantitative analysis of points is shown in Figure 12. Typical X-ray dispersion micro-analysis (EDX). (atomic %).

Spectrum	%C	%N	%V	%Cr	%Fe	%Co	%Mo	%W
1	54.05	14.64	11.45	2.89	16.97	--	--	--
2	55.38	20.44	8.68	3.15	12.35	--	--	--
3	--	25.67	2.9	3.32	63.64	2.23	1.39	0.85
4	55.54	10.42	13.44	2.68	17.92	--	--	--
5	54.54	15.49	16.98	2.03	10.96	--	--	--
6	--	24.29	2.44	4.65	64.42	1.89	1.35	0.96

shows the results of the semi-quantitative analysis of carbides and matrix constituents in the points specified in Figure 12. It may be observed that carbides are affected by nitriding, reaching variable levels which are higher than 10% in N (atomic %). Nevertheless, it seems that the conclusion is that in the matrix constituent, the values reached are higher than 20%. Anyway, and it may be observed in Figure 11, N content seems to be influenced by the depth reached in the nitrided layer.

4. Conclusions

In this work, it has been studied the effect of nitriding and shot peening surface treatments on wear resistance and impact toughness of the powder metallurgy steel Böhler K390. For this purpose, several factors related to this steel quenching and tempering have been previously modified. The following conclusions are the most important ones:

• Before implementing surface treatment, it is recommended to carry out austenitizing treatment at 1100 °C with air cooling and 3 tempering treatments at 550 °C. This treatment gives the possibility to have less retained austenite amount (~3%), more carbide density (~15–16%) and the highest level of impact toughness without a notch, above 40 J/cm². Besides, if this treatment is combined with further ion nitriding, a really high level of abrasive wear resistance will be obtained.
• The only precipitated carbide is of the MC type, associated with V. These carbides also show the presence of Cr and Mo. These latter elements, although also found in solid solution, appear in greater proportion in these carbides. Co and W tend to remain in solid solution.
• Both the shot penning treatment as well as ion nitriding offer a considerable increase in hardness, reaching values of 1500–1600 HV. Nevertheless, the highest level of abrasive resistance is obtained after ion nitriding, verifying that shot peening treatment does not improve this resistance.
• The nitrided layer thickness is between 150 and 175 μm, and the N content gradually decreases from the surface towards the inner part. Carbides are affected by nitriding and, at an intermediate depth of the nitrided layers, they have about 10% atomic values in N variables. These values are higher in the matrix constituent, even above the 20% atomic value in N.