Modelling approach to support design of new high nitrogen austenitic steel for wind turbines components

Abstract

The alloy design of a new high nitrogen austenitic steel for bearing used in wind turbine gearboxes has been approached. The alloy design was based on a modelling approach to the different aspects of manufacturing routes. The mechanical strength was predicted by means of empirical models. Nitrogen solubility in liquid steel was estimated on the basis of Sievert’s law and Thermo-Calc calculations. Scheil calculations and models developed by Centro Sviluppo Materiali SpA (CSM) were used for estimating nitrogen solubility. The stability of austenite has been studied by thermodynamic calculations. The results show that amounts of at least N = 0.5% are required to fall in austenite domain approaching the required mechanical strength. Vacuum induction melting + pressure electroslag remelting is required to manufacture alloys with N ⩾ 0.6%, but the addition of elements as Mn can potentially allow to increase N amounts achievable with only vacuum induction melting route processing.

Keywords

High nitrogen austenitic steel alloy design vacuum induction melting pressure electroslag remelting wind turbine gearbox bearing

Introduction

Among renewable energy sources, the wind energy sector has grown significantly within the last two decades. One of the critical points is the gearboxes life expectancy – particularly in offshore applications – with the related costs of the failures. Accordingly, the increase in the reliability by improving the tribological, fatigue, surface fatigue and corrosion properties of steel bearings, represents a very important issue. High carbon steel known as 100Cr6 (52100) is the most widely used in bearings production. Innovative steel grades can be applied in gearboxes of wind turbines with the aim of increasing the service lifetime and therefore reducing costs. High nitrogen austenitic steel (HNAS) can improve the mentioned critical properties of the bearings: nitrogen alloying can enhance fatigue strength, wear and fatigue resistance as well as resistance to crevice corrosion and pitting corrosion with respect to standard austenitic stainless steels. Furthermore, nitrogen addition helps to increase mechanical strength and can be used instead of nickel as an austenite-forming.

In this work, a new HNAS has been designed for bearings used in wind turbine gearboxes, considering two different process routes: (1) vacuum induction melting (VIM) in N₂ atmosphere and (2) VIM in N₂ atmosphere followed by pressure electroslag remelting (VIM + PESR). The alloy design was carried out by means of a modelling approach taking into account the different aspects of manufacturing routes.

Methods

On the basis of the specifications of bearing manufacturers, a minimum level of ultimate tensile strength (UTS) ~ 1000 MPa (30 HRC) has been considered for the bulk, while the strength in a superficial layer of about 3 mm needs to be raised up to yield strength (YS) ~ 1500 MPa and UTS ~ 2000 MPa (58 HRC) through work hardening. As an example, with a nitrogen content of 0.6% and a cold deformation rate of 40/50% a YS above 1500 MPa can be achieved (Paulus et al., 1993). In high nitrogen bearing austenitic steels, nitrogen present as interstitial solute atoms without making any nitrides has a very effective strengthening effect due to solution hardening, higher than that of carbon (Irvine et al., 1961). The mechanical strength of the different compositions studied has been evaluated by means of empirical relationships, summarized in Table 1 (Irvine et al., 1969; Ruffini, 2005; Sanchez et al., 1999; Speidel, 2006).

Table 1.

Empirical relationships for prediction of mechanical properties.

Author	Relationships	Notes/ranges of validity (% wt.)
Irvine et al. (1969)	YS = 63.5 + 496N + 356.5C + 20.1Si + 3.7Cr + 14.6Mo + 18.6V + 4.5W + 40.3Nb + 26.3Ti + 12.7Al + 2.5δ + 7.1d^−1.2 UTS = 449.5 + 852.5N + 542.5C + 37.2Si − 1.5Ni + 18.6Mo + 77.5Nb + 46.5Ti + 18.6Al + 2.2δ + 7.1t^−1.2	δ is the content of δ-ferrite in vol.%, d is the grain size in m and t is the number of twins per mm
Speidel (2006)	YS = 150 + 500N^−1.2 UTS = 500 + 500N^−1.2
Sanchez et al. (1999)	YS = 36.7 + 533.7N + 35.5Si + 17.1Mo + 11.2Cr + 3.2Mn − 3.3Cu − 3.7NiUTS = 336.9 + 436N + 38.9Si + 25.9Mo + 12.4Cr + 3.1Mn − 2.8Cu − 4.9Ni	C ⩽ 0.13, Si ⩽ 0.15, Mn = 5.0–14.5, Cr = 14–20, Ni = 0.5–8, Mo ⩽ 0.3, Cu = 0.5–4, N = 0.08–0.54
Ruffini (2005)	YS = 325.1 + 356.79N − 3.26MnUTS = 591.9 + 272.47N + 4.12Mn	C = 0.06, Si = 0.35, Mn = 4–12, Cr = 18.1, Ni = 3, S < 0.03, Cu = 2.1, N = 0.15–0.55

The chemical composition needs to be balanced aiming at having a temperature range available for full stability of austenite phase (free of chromium-rich carbides, nitrides and residual δ-ferrite) to perform thermomechanical and solubilization treatments, and maximizing nitrogen solubility in liquid phase and during solidification. Austenite stability at room temperature is also important since δ-ferrite is detrimental for localized corrosion resistance and hot workability. Designed alloys have been evaluated relatively to their position respect to borderline of γ/γ + δ domains at room temperature in modified Schaeffler diagrams, specifically tuned for high nitrogen steels (Rechsteiner and Speidel, 1993; Uggowitzer et al., 1996), summarized in Table 2. Candidate alloys should fall in γ domain or close to (δ + γ) boundary, which is represented by the line of equation Ni_eq = k Cr_eq + d.

Table 2.

Modified Schaeffler diagrams.

Author	Relationships	Ranges of validity (% wt.)
Uggowitzer et al. (1996)	K = 1.148, d = −11.1 Cr_eq = Cr + 1.5Mo + 1.5W + 0.48Si + 2.3V + 1.75Nb + 2.5AlNi_eq = Ni + Co + 0.1Mn − 0.01Mn² + 18N + 30C	Cr = 15–18, Mo = 3–6, Mn = 10–12,N = about 0.9, Ni-free
Sanchez et al. (1999)	k = 1.309 d = −12.8Cr_eq = Cr + 0.7Si + 1.25Mo + 0.05MnNi_eq = Ni + 27.4C + 22.7N + 0.35Cu	C ⩽ 0.13, Si ⩽ 1.5, Mn = 5–14.5, Cr = 14–20, Ni = 0.5–8, Mo ⩽ 3, Cu = 0.5–4, N = 0.08–0.54
Rechsteiner and Speidel (1993)	k = 1.2 d = −12.12Cr_eq = Cr + 1.5Mo + 0.5Si + 2.3V + 2.5AlNi_eq = Ni + 0.12Mn − 0.0086Mn² + 30C + 18N + 0.44Cu	Cr = 18–25, Mn ⩽ 25, N = 0.6–1.1, Mo ⩽ 7

Since a wide domain of austenite stability is needed for thermomechanical and solubilization treatments, to avoid precipitation of chromium-rich carbides and nitrides, austenite stability has also been evaluated at high temperatures by means of thermodynamics calculations. An example (JmatPro software) is reported in Figure 1.

Figure 1.

Example of calculation of austenite stability (JmatPro).

To estimate nitrogen solubility in the liquid phase, different empirical relationships have been considered (Medovar et al., 1996; Satir-Kolorz, 1990; Satir-Kolorz et al., 1989; Schurmann and Kunze, 1967) derived by Sievert’s law

N_{limit} = \frac{\sqrt{p_{N_{2}}^{1 / 2}}}{k f_{N}}

(1)

where k is the constant of equilibrium for nitrogen dissolution in the liquid melt, pN₂ is the partial pressure of nitrogen and f_N is the Henrian activity coefficient of nitrogen. Nitrogen solubility decreases with temperature in the liquid melt and is enhanced by high pN₂ values, which can be achieved, for instance, in a PESR process. Such methods were applied to assess nitrogen solubility limit to evaluate the risk of gas formation in the liquid melt. The results of estimations were also compared with evaluations made by Thermo-Calc code.

The gas formation during solidification has been studied by means of a physical model developed at CSM, which allows to estimate the risk for gas nucleation due to nitrogen microsegregation, taking into account of the effect of back-diffusion in the solid phase, as well as cooling rate and the pressure (Ridolfi and Tassa, 2003).

Alloy design

Chemical composition range

The ranges for nitrogen and manganese were selected as follows: N = 0.3–0.8%; Mn = 1–23%, in order to achieve an austenitic microstructure. Cr content was set at 18% in order to guarantee a high resistance to salty environment, comparable to Cronidur^® 30 (a high-performance high nitrogen martensitic steel). C content was set at 0.05% in order to avoid massive Cr-rich carbide precipitation with decay of intergranular corrosion resistance and embrittling effect. The amounts of the other elements were set considering them as typical impurities in industrial raw materials. Ni and Cu up to 2% were also considered. With such premises, the set of compositions of Table 3 was used for theoretical investigations.

Table 3.

Compositions selected for the theoretical investigations.

Alloy	C	Cr	Cu	Mn	Ni	N	Si	Mo	V
SW1	0.05	18.0	0.2	11.0	0.2	0.30	0.3	0.2	0.1
SW2	0.05	18.0	0.2	11.0	0.2	0.60	0.3	0.2	0.1
SW3	0.05	18.0	0.2	11.0	0.2	0.80	0.3	0.2	0.1
SW4	0.05	18.0	0.2	15.0	0.2	0.30	0.3	0.2	0.1
SW5	0.05	18.0	0.2	15.0	0.2	0.45	0.3	0.2	0.1
SW6	0.05	18.0	0.2	15.0	0.2	0.60	0.3	0.2	0.1
SW7	0,05	18.0	0.2	18.0	0.2	0.30	0.3	0.2	0.1
SW8	0.05	18.0	0.2	18.0	0.2	0.60	0.3	0.2	0.1
SW9	0.05	18.0	2.0	11.0	2.0	0.30	0.3	0.2	0.1
SW10	0.05	18.0	2.0	15.0	2.0	0.30	0.3	0.2	0.1
SW11	0.05	18.0	2.0	18.0	2.0	0.30	0.3	0.2	0.1
SW12	0.05	18.0	0.2	23.0	0.2	0.60	0.3	0.2	0.1
SW13	0.05	18.0	1.0	23.0	1.0	0.60	0.3	0.2	0.1
SW14	0.05	18.0	2.0	23.0	2.0	0.60	0.3	0.2	0.1
SW15	0.05	18.0	1.0	23.0	1.0	0.45	0.3	0.2	0.1

Mechanical strength

According to predictions by the empirical relationships (Table 1), the most promising compositions are those containing a nitrogen content of at least 0.6%. In particular, values of UTS approaching 1000 MPa can be expected.

Austenite stability

According to the modified Schaeffler diagram, a fully austenitic microstructure at room temperature can be obtained for the compositions with N ⩾ 0.6%. An example is shown in Figure 2.

Figure 2.

Modified Schaeffler diagram according to Speidel (2006) and positioning of selected alloys (Mn = 11–18%).

As for as the austenite stability at high temperature is concerned, the estimations according to JmatPro and Thermo-Calc generally agree and the following considerations can be made:

Nitrogen widens austenite stability, pushing temperature of γ → (γ + α) transformation towards higher values; at the same time, temperature of nitrides precipitation increases with nitrogen content.

Manganese at elevated concentrations stabilizes ferrite and reduces temperature of nitride precipitation.

Copper and nickel stabilize austenite respect to δ-ferrite transformation and increase the temperature of nitrides precipitation.

Nitrogen solubility

Empirical models used to estimate nitrogen solubility generally agree in estimating absence of gas nucleation at 1600°C and atmospheric pressure only for alloys containing lower nitrogen contents (0.3%). Achievement of higher nitrogen contents as 0.45% appears critical adopting manufacturing routes at atmospheric pressure. Further evaluations have been made by means of Thermo-Calc, at different pressures (Figures 3 and 4). Increasing pressure up to 10 atm would allow solubilization of N contents up to 0.8%. At atmospheric pressure, nitrogen solubility increases by increasing Mn content, by decreasing temperature and decreases by increasing Cu and Ni content. Solubilization of N at 1600°C could be feasible for Mn content of 23% only for compositions containing N up to 0.45%. A reduction of temperature from 1600°C to 1500°C would allow solubilization of N = 0.6% for Mn = 23% at conditions close to atmospheric pressure.

Figure 3.

Nitrogen solubility according to Thermo-Calc. Alloys with Mn = 11–18%.

Figure 4.

Nitrogen solubility according to Thermo-Calc. Alloys with Mn = 23%.

During solidification, nitrogen and other solutes partition occurs from growing solid phase towards residual liquid melt, thus resulting in an increase in solute content at the final stages of solidification range. Preliminary estimations for gas nucleation by means of the physical model by CSM have been performed. An example referring to composition SW12 is reported in Figure 5, at conditions lower than atmospheric pressure (0.9 bar) and T = 1500°C, which could be performed in standard VIM route. The conditions for gas nucleation due to nitrogen segregation during solidification are not satisfied.

Figure 5.

Estimation of gas nucleation during solidification at pressure condition lower than p_atm for alloy SW12 at p(N₂) = 0.9 bar and T = 1500°C.

Conclusion

The methods used to design a new class of HNAS steels have been described. Compositions can be defined using different models to meet mechanical strength targets and produce gas pore-free materials. Austenitic steels strengthened by N solid solution are expected to have high mechanical resistance. Strength has been evaluated by empirical relationships influenced by chemical composition, with particularly emphasis on N content. Combining the different models used to establish stability of austenitic phase, the following considerations can be drawn:

Compositions with N = 0.3% do not fulfil technical specifications due to low mechanical resistance in the annealed state and poor stability of, even in case of alloying with Ni and Cu.

Compositions with N ⩾ 0.6% fall in γ domain and show a limited stability of δ-ferrite which disappears for N = 0.8%. Compositions with N = 0.45% show an intermediate behaviour and fall near to boundary between γ and (γ + α) domains. Mn stabilizes δ-ferrite and lowers temperature for Cr₂N precipitation, increased by N, Ni and Cu addition.

Compositions with N = 0.6% can be manufactured via PESR route and potentially at atmospheric pressure for high Mn content (23%) even according to experiences reported in literature (Saller et al., 2004; Svyazhin et al., 2014). Processing of some alloys (SW12, SW15) could be feasible if manufacturing conditions are adjusted respect to nitrogen content: N = 0.60% could be reached at 1500°C and pN₂ = 0.9 bar, N = 0.45% could be reached at 1500°C and pN₂ ⩾ 0.6 bar.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research leading to these results has received funding from the European Union’s Research Fund for Coal and Steel (RFCS) research programme under grant agreement no RFSR-CT-20014-00018.

References

Irvine

Gladman

Pickering

(1969) The strength of austenitic stainless steels. Journal of the Iron and Steel Institute 207: 1017–1028.

Irvine

Llewellyn

Pickering

(1961) High-strength austenitic stainless steels. Journal of the Iron and Steel Institute 199: 153–175.

Medovar

Ya Saenko

Gigorienko

. (1996) Arc-Slag Remelting of Steels and Alloys. Berlin: Cambridge International Science Publishing Ltd, p. 36.

Paulus N, Uggowitzer PJ, Müllner P, et al. (1993) Cold and warm work of austenitic nitrogen steels. In: Proceedings of innovation stainless steel, Florence, 11–14 October 1993.

Rechsteiner

Speidel

(1993) New methods for the production of high nitrogen stainless steels. In: Proceedings of the innovation stainless steel, Florence, 11–14 October 1993, pp. 2107–2111.

Ridolfi

Tassa

(2003) Formation of nitrogen bubbles during the solidification of 16–18% Cr high nitrogen austenitic stainless. Intermetallics 11: 1335–1338.

Ruffini

(2005) Solidificazione degli acciai inossidabili austenitici a basso tenore di Nickel. PhD Thesis, Università di Roma ‘La Sapienza’, Rome.

Saller

Bernauer

Leitner

. (2004) In: Proceedings of the 7th international conference on high nitrogen steels (HNS) 2004, vol. 2, pp. 283–292. Available at: http://www.gbv.de/dms/tib-ub-hannover/470166142.pdf

Sanchez

Fernandez

del Campo

. (1999) Non-standard austenitic stainless steels, with high nitrogen content: Influence of alloying elements on phase equilibrium (γ-δ) during solidification and heat treatments. In: Proceedings of the stainless steel ‘99 science and market, Chia Laguna Sardinia, 6–9 June, vol. 2, pp. 353–360.

10.

Satir-Kolorz

(1990) Concept of high nitrogen steel. In: Proceedings of the 3rd international conference on high nitrogen steels: HNS 93, Kiev, 1993. Available at: http://www.gbv.de/dms/tib-ub-hannover/19452194X.pdf (also reported in HK Feichtinger).

11.

Satir-Kolorz

Feichtinger

Speidel

(1989) Ǘber die Löslichkeit von Stickstoff in Eisen-und Stahlgußlegierungen unter erhöhtern Druck. Giessereiforschung 41(4): 149–164.

12.

Schurmann

Kunze

H-D

(1967) Arch. Eisenhuettenw. 38: 685–689.

13.

Speidel

(2006) Nitrogen containing austenitic stainless steels. Materialwissenschaft und Werkstofftechnik 37(10): 875–880.

14.

Svyazhin

Kaputkina

Bazhenov

. (2014) Formation of gas bubbles during solidification of nitrogen-alloyed stainless steel. In: Proceedings of the international conference: HNS 2014, Hamburg, 16–19 September, pp. 20–26. Essen: Energietechnik Essen.

15.

Uggowitzer

Magdowski

Speidel

(1996) Nickel free high nitrogen austenitic steels. ISIJ International 36(7): 901–908.