Clean and Sustainable Hydrogen-Electric Propulsion

*Xin Gao and Chengwei Zhao*

## **Abstract**

For future hypersonic and supersonic flight, clean, sustainable and energy-efficient propulsion should be addressed in the general background of the sensational clean electric transition of aircraft. This chapter is to draw the attention of the research communities on the possible feasibilities and challenges of hydrogenelectric propulsion in hypersonic and supersonic flight. This chapter is structured with the following aspects, (1) general design and hybridisation concepts of hydrogenelectric propulsion for general aircraft and their hypersonic and supersonic considerations; (2) merits of hydrogen-electric propulsion on thermofluids process integrations; (3) potential merits of hydrogen-electric propulsion projected through thermofluids structural engineering and re-engineering; (4) storage options and their challenges in design and operation; and (5) reliability considerations.

**Keywords:** hydrogen, cryogenic, electric propulsion, fuel cells, system design, operation

## **1. Introduction**

Hydrogen has been considered a very promising clean, sustainable energy source for a long time. In recent decades, it has come back into the limelight [1]. Due to the current world situation, accompanied by changes in the climate and the increase in the price of traditional fossil energy sources, aircraft urgently needs a new propulsion technology to reduce polluting emissions and lower fossil fuel consumption [2]. Hydrogen propulsion is, therefore, widely seen as a solution to the current situation because of its renewable nature and the effective reduction of carbon dioxide emissions [3]. There are currently four mainstream hydrogen propulsion: (1) all-electric fuel cell propulsion, (2) fuel cell propulsion and complementary direct-drive hydrogen turbines, (3) hydrogen-fuelled turbofan electric propulsion and (4) pure combustion propulsion [4]. The mission range of the aircraft still determines the efficiency of these concepts. In the medium to long range, fuel cells will be more advantageous [5]. However, fuel cells still face more challenges in being used as an aviation application. Because of its lower power-to-weight ratio, the system's complexity [6] and the inconvenience of hydrogen storage and waste heat management [7] are all issues that need to be addressed. However, current trends suggest that aircraft capable of full hydrogen fuel cell propulsion remains the ultimate goal of long-term research. The

challenges related to the application of hydrogen fuel cells in aviation are described in detail in Section 3, along with current research and future perspectives.

## **2. Fuel cell principles**

A fuel cell is an energy conversion device that converts chemical energy stored in fuels and oxidisers into electrical energy through a redox reaction based on electrochemical principles. The fuel cell itself does not store energy; it is, like the internal combustion engine, a device that converts chemical energy into other forms of energy using 'fuel'. An internal combustion engine produces heat by burning fuel, and thermodynamic energy is converted into kinetic energy. On the other hand, a fuel cell has electrical power through an electrochemical reaction, and the electrical energy produced can be converted into other energy as required, such as mechanical energy through an electric motor. In this respect, it is more like a battery with a fuel tank. However, unlike conventional batteries, it does not require charging time, only refuelling, and it could have a higher energy density than traditional batteries especially on the system level. And the device itself produces no noise or vibrations in the workshop, and this electrochemical reaction does not produce pollutants that are harmful to the climate [8].

## **2.1 Working principle**

The fuel cell consists of four main components: the anode, the cathode, the electrolyte, and the external circuit. Fuel gas and oxidation gas are fed through the anode and cathode of the fuel cell separately. The fuel gas emits electrons at the anode, which are transferred to the cathode via an external circuit and combined with the oxidation gas to form ions. The ions migrate through the electrolyte to the anode under the influence of an electric field and react with the fuel gas, forming a circuit and generating an electric current. At the same time, the fuel cell generates some heat due to its electrochemical reaction and the cell's internal resistance. In addition to conducting electrons, the cathode and anode of the cell act as catalysts for redox reactions (**Figure 1**).

#### **2.2 Fuel cell types**

There are numerous classifications of fuel cells, commonly based on the type of electrolyte, including proton exchange membrane fuel cells (PEMFCs), alkaline fuel cells (AFCs), phosphoric acid fuel cells (PAFCs), molten carbonate fuel cells (MCFCs), and solid oxide fuel cells (SOFCs) [9]. In addition, different variants of these types have been developed, such as the direct methanol fuel cell (DMFC), which is based on LT-PEMFC but uses methanol and the high-temperature (HT-)PEMFC, in which the phosphoric acid is stabilised in a polymer membrane. Proton ceramic fuel cells (PCFCs) have many similarities to SOFC but use proton-conducting electrolyte materials [10].

The Proton Exchange Membrane Fuel Cell (PEMFC), also known as the Polymer Electrolyte Membrane Fuel Cell, was invented by General Electric in the late 1950s and used in NASA space missions [11]. The electrolyte is a very thin polymer membrane in this type of cell. This polymer membrane conducts protons but not electrons, thus ensuring ion exchange between the electrodes. Generally, proton exchange membrane fuel cells use platinum on carbon (Pt/C) as the catalyst for the cathode reaction (**Figure 2**).

The different membrane materials determine the specific conditions of use. LT-PEMFCs operate between 65°C and 85°C, which offers the possibility of using water *Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

**Figure 1.** *Simplified schematic of the working principle of a fuel cell.*

**Figure 2.** *Working principle of PEMFC.*

cooling in the fuel cell cooling system. HT-PEMFCs operate between 140°C and 180° C, where the proton conductivity is high enough, and the polymer membrane remains chemically stable. Ceramic membranes used in SOFCs require temperatures of 500– 1000°C to achieve sufficiently high ionic conductivity, depending on the electrolyte material and thickness. These operating temperatures determine the materials of the cell and system.

## **2.3 Fuel cell systems**

As the chemical electromotive force of a fuel cell is well below 1 V, the clamping voltage and the electrical power generated can be increased by connecting several cells in series to make the fuel cell fit for everyday use. Each cell is separated by interconnected bipolar plates and sealed with gaskets, thus forming a fuel cell stack. By designing the bipolar plate structure, air and fuel can be evenly distributed within the cell, thus increasing the efficiency of the fuel cell. For low-temperature fuel cells where liquid cooling can be used, the passage of the coolant is also an essential part of the bipolar plate structure. (This is described in more detail in Section 3.2).

The fuel cell stack is the core part of the overall fuel cell system and is used to generate electricity. At the same time, to ensure the fuel cell stack's stability, other components must work together to supply the fuel, air, and coolant to the fuel cell stack in the correct operating environment. Depending on the operating environment, different parts are used, often referred to as BoP (Balance of Plant). These may include pumps, blowers, valves, heat exchangers, humidifiers, filters, chemical reactors, injectors, burners, gas purification, electric actuators, power converters, and everything else required for the operation of the fuel cell system.

### **2.4 Fuel cell characteristics**

Different types of fuel cells have different physical characteristics, and it is crucial to select a suitable type of fuel cell for the best possible application in various fields. In 2009, the German Aerospace Centre (DLR) used a 16 kW high-temperature polymer electrolyte membrane fuel cell (HT-PEMFC) in an Antares motor glider for a test flight [12]. This system was upgraded with a 30 kW low-temperature polymer electrolyte exchange membrane fuel cell (LT-PEMFC). The experience from the appeal platform led to another successful human-crewed test flight in 2016 [13, 14]. Based on the application of fuel cells in aviation, the following section focuses on the characteristics of low-temperature polymer electrolyte membrane fuel cells (LT-PEMFC) and high-temperature polymer electrolyte membrane fuel cells (HT-PEMFC).

## *2.4.1 Lt-PEMFC*

LT-PEMFC typically operates at temperatures between 65°C and 85°C. Its operating temperature is limited because it uses a solid polymer-acid membrane (usually PFSA) which conducts protons when wetted. The character of this hydrated membrane is such that the LT-PEMFC must be operated at a temperature below the boiling point of water. However, the operating temperature must be neither too low (below 65°C) to flood the polymer membrane with condensed water nor too high (above 85°C) to degrade the polymer membrane after dehydration [15]. In addition to the control of temperature, the management of water inside the fuel cell is also a

## *Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

crucial point. As the product of the hydrogen fuel cell reaction is water, if the water produced through the electrochemical reaction is not managed, too much water will soak into the electrodes, and the reactants will not be able to enter the reaction point. Dehydration of the polymer membrane will reduce ionic conductivity, so proper wetting of the polymer membrane is also required. The wetting of the membrane can be maintained by the water produced in the electrochemical reaction, and the excess water must be removed. To effectively control moisture in the fuel cell, individual gas diffusion layers can be designed into the cell structure to allow the gas reactants to participate more efficiently in the reaction and to direct the drainage of liquid water from the electrodes [16]. In addition, catalysts are required to increase the power density of the fuel cell to make it more efficient. In general, platinum is a perfect catalyst for electrochemical reactions, which can effectively improve the efficiency of fuel cells at low temperatures.

The LT-PEMFC system requires external air to participate in the reaction within the fuel cell, and an air compressor and air filter are usually used to supply air to the positive electrode. The air must be humidified before entering the positive electrode for reaction, either externally using a humidifier or internally within the fuel cell. The most common fuel for LT-PEMFC is hydrogen, which must be supplied at the correct pressure and temperature. For the reactions within the fuel cell to be more efficient, the hydrogen needs to be evenly distributed and adequately humidified. LT-PEMFCs are less tolerant to impurities contained in the fuel. This is mainly due to their relatively low operating temperature, which results in the solid adsorption of impurities on the platinum catalyst at the cathode. To prevent the accumulation of pollutants or contaminants in the fuel at the anode compartment, the fuel needs to be cleaned, which results in a slight loss of fuel (<1%) [17]. Hydrogen produced from natural gas carries sulphur and carbon monoxide, which can seriously affect the lifetime and performance of the LT-PEMF. Carbon dioxide can also have an adverse effect through the formation of carbon monoxide in the side reactions. Ammonia contamination can also lead to rapid degradation as well as membrane poisoning [18]. Most LT-PEMFC systems today use high-purity hydrogen to achieve a satisfactory lifetime with minimal loading of the platinum catalyst. In principle, however, LT-PEMFC can be operated on modified hydrocarbon fuels and cracked ammonia, provided that the concentrations of sulphur, ammonia and carbon monoxide are all well below 1 ppm. Fuel cells with fuel handling systems have been built and demonstrated [19].

Bipolar plates are another vital part of the LT-PEMFC system and integrate many functions. Bipolar plates are usually designed with a complex hydrodynamic structure that allows them to perform the tasks of homogeneous distribution of fuel and air, separation of air and fuel, management of water and heat inside the fuel cell and conduction of current. Bipolar plates can be made of graphite, metal or composite materials [20]. The choice of bipolar plate material can significantly impact the system's cost, functionality, lifetime, weight and size; for example, metal plates have a higher energy density but a shorter lifetime than graphite. The correct bipolar plate material for each application and an excellent structural design are essential.

The heat management of the LT-PEMFC is also an important aspect. Usually, a liquid cooling system is generally used, where the coolant is a mixture of water and anti-freeze additives. The liquid cooling system usually consists of a coolant pump and a radiator. To prevent ions in the coolant from causing leakage currents in the battery stack, a Raisin filter should also be installed to filter out the ions in the coolant.

### *2.4.2 HT-PEMFC*

To solve the problems of LT-PEMFC, the first attempts have been made to increase the operating temperature of PEMFCs. Therefore, it is vital to change the performance of proton exchange membranes so that they can be used in a broader range of temperatures to meet the requirements of both low-temperature cold start and high performance after start-up, which will become an important direction in the future, that is, high-temperature proton exchange membrane fuel cells (HT-PEMFC).

High-temperature proton exchange membrane fuel cells (HT-PEMFCs), which operate at 100–200°C, have more key advantages than ordinary LT-PEMFCs: (1) the catalytic activity of the electrodes is highly active; (2) the catalyst has high resistance to impurity gases; (3) there is only gas phase mass transfer, which is simple and efficient [21]; (4) the hydrothermal management is simple; (5) the methanol permeation problem of methanol fuel cells is solved at high temperatures; (6) there is no need to consider separation and purification devices, and the reforming gas can be matched with online hydrogen production [17], making it possible to produce inexpensive liquid reforming hydrogen (e.g. methanol and acetic acid), and it is easy to store, transport and refill, etc.

As a core component of HT-PEMFC, the proton exchange membrane has an important impact on the lifetime and performance of the fuel cell, and the development of HT-PEM is of very positive significance. However, the technology has some unavoidable challenges: ① The proton conductivity of existing membranes at high temperatures in the temperature range of 100 200°C is severely reduced, and the perfluorosulphonic acid proton exchange membranes currently in use are heavily dependent on the water content of At high temperatures, the water content in the membrane is low, and the proton carriage mechanism is severely weakened [22]; ② the stability of existing membranes at high temperatures is poor; ③ the performance degradation of fuel cells at high temperatures such as carbon corrosion and platinum dissolution will be exacerbated and durability and lifetime will be reduced [23]; ④ how to solve the compatibility between high-temperature membranes and catalysts [21].

LT-PEMFC systems are already widely used in many areas such as automotive, submarines, portable power supplies, specialist UAV areas and aircraft propulsion. Low operating temperatures offer many advantages, such as lower manufacturing costs, longer service life, high reliability and short start-up times. However, the design of the heat management system of the LT-PEMFC is also a significant challenge due to the operating temperature requirements. The energy-to-weight ratio of the LT-PEMFC is relatively low compared with fuel cell systems operating at higher temperatures. Still, the lower operating temperature and the materials used result in better service life and lower manufacturing costs.

## **3. Aviation PEM fuel cells**

#### **3.1 Aviation challenges**

Civil aviation has always held a relatively large share of transport, especially for medium- and long-haul passenger transport, where the aviation industry holds a significant market share and is growing at a rate of around 3–4% per year. It also brings the problem of contributing to a large amount of greenhouse gas emissions, such as carbon dioxide. According to statistics today, the aviation industry accounts for 3% of global CO2 emissions; by 2050, this will be 24% [4]. So how to reduce greenhouse gas emissions and achieve the goal of green flight has become an urgent issue for the aviation industry.

Hydrogen fuel cell systems are widely considered a new type of aviation propulsion system that can replace conventional fuel engines in the future. With its lower noise level, clean, sustainable energy (hydrogen) use and more efficient energy conversion, it is perfect for achieving the goal of truly green flight.

However, there are still many hurdles to overcome before hydrogen fuel cell systems can truly replace conventional fuel engines as the primary propulsion system in large civil aircraft. Firstly, to meet the power requirements of large passenger aircraft, fuel cell systems need to have an output of megawatts or even tens of megawatts. Take the DLR study as an example, a fuel-cell-powered four-seater aircraft (HY4) flown by DLR in 2016. This prototype was equipped with a hybrid power system consisting of an LT-PEMFC and a lithium-ion battery, with the LT-PEMFC acting as the primary output for the main power and the lithium-ion battery as an auxiliary for the aircraft during take-off and climb. The power system has an output of 80 kW [24], and the successful test flight of this aircraft represents the possibility of hydrogen fuel cells as a propulsion system for small manned aircraft. However, for application on large passenger aircraft, a large number of fuel cell units would need to be connected in series to achieve the required power.

To ensure the proper functioning of many fuel cell stacks connected in series, extremely high demands are placed on all system parts. The first is the thermal management of the entire system. As the LT-PEMFC is sensitive to the operating temperature requirements, an efficient and stable thermal management system is necessary. The thermal management of a fuel cell system of this power level is also a severe challenge. Due to the characteristics of the proton exchange membrane used in the LT-PEMFC, a water management system is also required to cope with a large number of cells in the fuel cell system to prevent each fuel cell from being dehydrated and flooded, causing power reduction or even failure, to ensure the regular operation of the whole system.

In aircraft design, weight is always a priority. The overall mass of the aircraft affects the energy consumption; the heavier the mass, the more fuel is required, and the more powerful the propulsion system needs to be to overcome the drag in flight and meet the required lift. This is why the lightweight design of fuel cell systems is also a challenge (**Figure 3**).

There are two ideas for reducing the weight of a fuel cell system. One is to reduce the number of fuel cell stacks, which means increasing the fuel cell unit's efficiency of the fuel cell unit, that is, increasing the power-to-weight ratio. The efficiency of the LT-PEMFC can be effectively improved by developing new types of proton exchange membranes, optimising the design of the flow field inside the fuel cell, etc. The second idea is to simplify and lighten the creation of the system, for example, by developing new fuel cell thermal management systems and using lightweight materials.

For large passenger aircraft, safety always comes first; each new aircraft must undergo rigorous certification before servicing. This requires a higher level of safety redundancy for essential systems. For fuel cell systems, increasing the redundancy of the fuel cell may significantly reduce the specific power of the entire system. A fuel cell system's specific power and lifetime are also considered essential factors for its application in aviation. While the specific power of fuel cells has been increasing over the last decade through the optimisation of components and manufacturing processes, the reliability required in aviation has been significantly less.

**Figure 3.** *The drag power of the hybrid PEMFC-jet.*

In this chapter, the increase in specific power for sizeable civil aircraft fuel cell systems was sought from a different approach rather than focusing on the design of the fuel cell system.

## **3.2 Cooling systems**

The cooling of a small LT-PEMFC can rely on air cooling, which is usually sufficient to keep the cell stack at a suitable temperature with the air involved in the reaction. However, when applied to megawatt-class LT-PEMFC systems, huge heat exchangers are required to maintain their operating temperature. This solution is not suitable for practical use. A large heat exchanger increases the air resistance of the aircraft in not only flight but also its weight is enormous, which adds additional drag to the plane. Liquid cooling systems are more widely used in automobiles due to their simple construction and better cooling capacity. But for aerospace applications, where the power of the system has been increased from kW to MW level, phase change cooling (PC) systems seem to have an advantage over liquid-cooled systems. Phase change cooling not only reduces weight but also increases cooling capacity by tens of percentages (**Figure 4**).

Therefore, a heat pump (HP) between the fuel cell system and the heat exchanger is proposed. According to Newton's law of cooling, the heat pump is used to increase the temperature of the coolant by the required heat exchanger capacity (surface area, volume and weight), that is, the flight drag can be effectively reduced. This PCHP (phase change–heat pump) cooling concept was first conceived for LT-PEMFC aeronautical applications in all published literature and the field of hydrogen propulsion in general, referring to Ref. [25].

To study and analyse the above-proposed PCHP cooling system, modelling and analysis were carried out based on a proven aircraft design and a typical flight mission.

#### **Figure 4.**

*Cooling methods for LT-PEMFC vs. power level.*

**Figure 5.** *Schematic presentation of the hybrid PEMFC-jet engine layout.*

A hybrid LT-PEMFC-jet engine layout was chosen and then fine-tuned in size under two different PEMFC cooling schemes. Key performance indicators, including two drag items, are extracted and compared. Finally, they are discussed and summarised.

The Airbus A320–200 jetliner was chosen for this study. The hybrid PEMFC-jet engine system consists of a fuel cell system and a hydrogen internal combustion engine (i.e., a hydrogen-fuelled turbine engine). **Figure 5** shows the entire layout.

To facilitate the study, the fuel cell system in the model consists of a fuel cell stack, a hydrogen tank and a heat exchanger. The hydrogen tank feeds the stored hydrogen to the fuel cell and the jet engine, which provides additional power during take-off by burning the hydrogen directly. The thermal management system and the fuel cell system compressor are not shown as they are included in the 'PEMFC System' block. **Figure 6** shows the propulsion power requirements for the different flight phases of a mission, and it can be seen that an enormous amount of power is required during the take-off phase. During the cruise phase, the operating environment is more stable, and propulsion is provided by the fuel cell. For the specific aeroplane, based on data from Kadyk et al [26]., the data also shows us the power requirements of the A320–200 at various stages, as detailed in **Table 1**.

**Figure 6.** *Analysed flight mission for hybrid PEMFC-jet engine.*


**Table 1.**

*Power demand of the A320–200 aircraft for take-off and cruising based on data from Kadyk et al. and resulting provided power by the jet engine and the fuel cell system.*

## *3.2.1 PEMFC cooling design*

A schematic diagram of the PCHP cooling circuit proposed in this work is shown in **Figure 6**. The cooling course (excluding the PEMFC stack) consists of a heat exchanger, a compressor (HP), an expansion valve, ideal piping and a coolant. Water has been chosen as the phase change coolant. For water to evaporate in the operating temperature range of the fuel cell stack, the coolant pressure must be lower than the

ambient pressure (0.2 bar). The low-pressure water (vapour mass of 0.153) evaporates, thus cooling the PEMFC stack and leaving a vapour mass of 0.950. Through the compressor, the pressure of the steam is increased by 5 bar, resulting in a significant temperature increase. Afterwards, the hot and pressurised steam (steam mass of 1.000) is transported to the heat exchanger, which removes heat into the ambient airflow around the aircraft. After passing through the heat exchanger, the steam is transformed into pressurised liquid water (vapour mass of 0.000). The heat exchanger model is implemented using the 'Condenser Evaporator (2P-MA)' module in Matlab/Simulink, and the minimum drag is determined using an optimisation algorithm. As the pressure drops through the expansion valve, the water is again mostly evaporated, that is, the vapour mass of the coolant returns to its original 0.153 at a pressure of 0.2 bar. Finally, the under-pressure cold water is piped into the PEMFC stack for a new cooling cycle. For comparison, we have also simulated conventional liquid cooling with the same cooling process as for the PCHP (**Figure 7**).

The reverse Rankine cycle of the coolant in the p-h diagram can be seen in **Figure 6**, where the blue line indicates the coolant cycle, the yellow line shows the isentropic process, and the red line indicates the isothermal process. The numbers 1–4 correspond to the states of the coolant in the PCHP cooling loop depicted in **Figure 8**.

## *3.2.2 PEMFC modelling*

The PEMFC is modelled to calculate the heat generated by the PEMFC system during operation and to analyse the cooling system. The following assumptions have been adopted to simulate the PEMFC subsystem of the engine.


#### **Figure 7.**

*a) the PCHP cooling loop proposed in this work for an aviation PEMFC system and b) schematic view of a conventional LC loop for the same PEMFC system.*

**Figure 8.**

*Heat pump cycle of the coolant inside the PCHP model. Blue lines: Coolant cycle; yellow lines: Isentropic process; red lines: Isothermal process. The numbers 1–4 correspond to the respective state of the coolant in the cooling loop.*

The subsystem-generated waste heat flow is [27],

$$
\dot{Q}\_{gen} = (V\_{th} - V\_{out}) \cdot i \cdot A\_{cell} \cdot N\_{cell} \tag{1}
$$

To remove the waste heat flow �Qgen to the ambient air, the two cooling designs mentioned above use the same aluminium louvre-fin heat exchanger and are of the same size [28].

#### *3.2.3 Drag power*

The drag generated by the PEMFC subsystem consists mainly of the system mass and aerodynamic friction caused by the heat exchanger, with some drag forces mitigated by the Meredith ramming effect [29]. This is because the air expands within the heat exchanger and leaves the nacelle faster than it entered, generating thrust. The overall drag is expressed as follows:

$$P\_{d,tot} = \upsilon\_{air,0} \times (D\_{d,max} + Dd, \text{zero}) + P\_{comp,r} + P\_{misc} - P\_{num} \tag{2}$$

where vair,0 is the free flow velocity of air (m/s); Dd, mass and Dd, aero are the mass and aerodynamic induced drag (N), respectively; Pcomp,r is the power consumption of the turbo compressor supplying fresh air from the environment to the fuel cell stack (W); Pram is the ram thrust caused by the Meredith ramjet effect (W). Pmisc is the miscellaneous drag power (W) depending on the other components in each cooling circuit. Dd; mass is determined by the force balance to lift resistance ratio L/D during the cruise. This ratio is the amount of lift generated divided by the aerodynamic drag due to movement through the air. The expression for the drag caused by mass is *Dd*,*mass* <sup>¼</sup> *msys*,*tot*�*<sup>g</sup> <sup>L</sup>=<sup>D</sup>* , where g is the gravitational constant (*m=s* 2); L/D is the lift-to-drag

ratio; and *msys*,*tot*, is the sum of all components in the system taken into account, which can be expressed as

$$m\_{\rm py,tot} = m\_{\rm HX} + m\_{\rm FC} + m\_{\rm JE} + m\_{\rm Tot,tank}, \\ m\_{\rm Tat,tank} = m\_{\rm Tank} + m\_{\rm H\_2} \tag{3}$$

The lift-to-drag ratio of an A320–200 airliner at cruise is [30]

$$L/D = 16.3$$

An analysis comparing the cooling effect and the dragging force of the powertrain of a PEMFC with two cooling systems shows that with a fixed cruise power requirement for the aircraft, all drag power items were found to rise monotonically with increasing current density based on simulations of the model. This is because the current density increase amplifies the heat flow through the system, amplifying the heat exchanger's capacity. This leads to an increase in absolute values for all resistance sources in the interval analysed. At higher current densities, the core drag and the drag caused by the heat exchanger mass rise more rapidly, while the ramjet thrust and external drag follow an almost linear trend. The relationship between the drag source and the size of the heat exchanger can explain this. While the external drag depends on the frontal area of the heat exchanger, the mass and core drag depend on the volume of the heat exchanger, which does not follow a linear trend. The total resistance of the liquid-cooled cooling loop and the phase-change heat pump cooling loop at different current densities is shown in **Figure 8** without accounting for the stack and hydrogen storage mass. It can be seen that the drag resistance first decreases and then increases through increasing current densities, indicating that there is a minimum drag resistance point for which the corresponding current density of the combustion link cell can be identified as the optimum operating point for the fuel cell. It can also be observed that the total drag force of the phase change heat pump cooling system is significantly lower than that of the liquid cooling system for both cold de-circuit designs. For the minimum total drag corresponding to the optimum current density, 0.5781 *A=cm*<sup>2</sup> for the liquid-cooled system and 0.7313 *A=cm*<sup>2</sup> for the phase-change heat pump system, it can be concluded that the phase-change heat pump system is advantageous in reducing the drag force during flight and increasing the efficiency of the fuel cell (**Figure 9**).

The combined performance of the hybrid system with its two different cooling systems is also worth discussing. The first thing to say is the total drag power of the

**Figure 9.** *a) the total drag of the LC loop. b) the total drag of the PCHP loop vs. stack current density.*

two systems: 6.648 MW for the LC system is 1.528 MW more than the PCHP cooling system, representing 15.79% of the aircraft's total propulsion power. As the coolant temperature in the PCHP system is much higher than in the LC system (324.4°C for the PCHP and 56.4°C for the LC), according to Newton's law of cooling and assuming other variables are inconvenient, the PCHP requires about 1/5 the area of the heat exchanger of the LC system. The use of a small-area heat exchanger also allows a further reduction in flight resistance and overall system mass. More drag requires more energy consumption, that is, more fuel needs to be carried to ensure range. At the same time, the current density of the LC system is also relatively low, which exacerbates the gap between the two systems.

Another point worth noting is the effect of the thrust generated by the Meredith Ramjet effect on the overall system for both systems. According to model calculations, the Ramjet effect of the LC cooling system provides an additional 1.773 MW of thrust, approximately three times that of the PCHP system. Still, it only counteracts 53.2% of the drag generated by the LC system, whereas the Ramjet effect in the PCHP cooling circuit counteracts 96.8% of the drag caused by itself. So the drag generated by the PCHP cooling system is entirely negligible, reducing the drag overcome by the propulsion system by 23%, which can also be seen as a 16% increase in airspeed.

## **3.3 Efficiency**

From the current research, it is clear that the performance of PEMFC itself still needs to be improved to be used in civil aviation. Previous sections have described the study of a sub-system of the PEMFC system, namely the cooling system. The performance of the entire hybrid PEMFC propulsion system has been optimised by employing a new phase change heat pump performance of the PEMFC [31]. To improve the performance of PEMFC, design optimisation of GDL is often considered an effective approach. Pore-scale simulations are generally regarded as one of the most effective tools for optimising the microscopic performance of porous media, such as GDL [32, 33]. In the last few decades, the rapid development of computer modelling and analysis, X-ray computed tomography and scanning electron microscope has provided a variety of effective modelling methods, simulating and studying GDLs at the microscopic level, such as pore-scale modelling (PSM), which is used below cooling system. This section will discuss how the performance of the PEMFC itself can be further optimised by design (**Figure 10**).

**Figure 10.** *Schematic of this multiscale modelling study approach.*

The Gas Diffusion Layer (GDL) is a critical component of the PEMFC structure and is responsible for managing water and heat and transporting substances in the PEMFC [34, 35]. The GDL transports the reactants to the catalyst layer (CL) and transports the products from the CL efficiently and uniformly, a link that impacts the performance of the PEMFC.

To design a new GDL, a 3D GDL model with a specific structure is first generated by the stochastic reconstruction method of the ANSYS Para-metric Design Language (APDL). The stochastic reconstruction method utilises a stochastic algorithm which allows the fibres to be placed randomly in a specified space. To facilitate the study, the following restrictions are imposed on the model: the generated GDL consists only of solid carbon fibres and pore scale, all carbon fibres on the same diameter and is entirely allocated on every single x-y plane, allowing for stacking between fibres. As a result, a numerical GDL is generated with only solid carbon fibres on the same diameter and is entirely allocated on every single x-y plane, allowing for the stacking of fibres. As a result, a numerical GDL is reconstructed, with a domain of size 304 <sup>304</sup> <sup>304</sup>*μm*<sup>3</sup> and a resolution of which the porosity is 0.78 and the fibre diameter is 8*μm*. **Figure 11 b**) analyses the pore size distribution to validate the reconstructed geometry. It demonstrates that the structure is ready for the GDL study.

ANSYS Workbench then carried out pre-processing to obtain the mesh coordinates. To improve its electrical and hydrothermal properties, the reconstructed GDL was rotated by different angles to form a new GDL. This step was transformed and tilted using MATLAB scripts and matrix transformation functions to obtain matrices with different degrees of GDL phase data. PSM, including gas diffusivity and electrical and thermal conductivities, extracted the effective transport properties of these angled GDLs. It should be noted that anisotropy is evident in common GDL materials. Calculating fluxes in the in-plane direction (x and y) and across the plane direction (z) is necessary. Finally, these properties were applied to a CFD model to investigate the performance of the PEMFC using this new design.

For ease of calculation, the simulation of the 3D macroscale model covers only one channel of the PEMFC, as shown in **Figure 12**. It consists of seven components: the anode current collector (ACC), the cathode current collector (CCC), the anode catalyst layer (ACL), the cathode catalyst layer (CCL), the anode gas diffusion layer (AGDL), the cathode gas diffusion layer (CGDL) and the membrane. Thermal and electrical conductivities at different GDL fibre angles were used in this model. Based on our PSM results, the effect of compression on the GDL has been considered.

In this macro model, the following assumptions are taken.

**Figure 11.** *Stochastic reconstruction of the GDL.*

### **Figure 12.**

*Schematics of the PEMFC model, 2-D view (left) and 3-D view (right).*


The conservation of mass governs this macroscopic PEMFC model, conservation of momentum, conservation of energy and electrochemical reaction equations [36–38].

With the help of the constructed micro and macro models, the properties of GDL, such as gas diffusivity, and electrical and thermal conductivity, are analysed.

The GDL was set at 15-degree intervals around the y-axis from 0 to 90 degrees, as shown in **Figure 13** and then rotated around the z-axis from 0 to 90 degrees. It can be observed that the porosities of the GDL appear to become more uniform as the angle increases. By calculating the effective gas diffusivity (EGD) using PSM, it was found that the effective gas diffusivity increased with increasing rotation angle.

To understand this phenomenon, relative gas diffusivity (RED) was introduced. As shown in **Figure 14a**, the RED increases with increasing angle. According to the

#### **Figure 13.**

*GDL angling, a) schematics of the angling around the y-axis and (b) porosity distributions through the z-axis direction of these angled GDLs.*

*Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

**Figure 14.** *Effects of the fibre angle on the EGD, (a) the EGD with the angle, (b) the RGD and relative porosity with the angle, including the calculation results and Bruggeman function.*

Bruggeman function, it can be concluded that EGD is correlated with porosities. This also shows that porosities are one of the determinants of EGD. However, we can also see from **Figure 14b** that the relative porosities are almost unity, which indicates that porosities' effect on EGD can be neglected. Research reveals that it is not porosity but tortuosity that causes EGD to increase with angle [39, 40]. Tortuosity is inversely proportional to EGD, with tortuosity decreasing as the angle increases, which is reflected in EGD increasing with the angle.

For carbon fibres, effective electrical conductivity (EEC) and effective thermal conductivity (ETC) increase with the angle but show an S-shaped curve as more electrons/heat is conducted through the fibres and less through the contact points between the fibres. This would result in less resistance to electrical/heat conduction through the planar direction [41, 42]. The difference between REEC and RETC is illustrated in **Figure 15** mainly because heat is also conducted through air, while electricity is only conducted through carbon fibres [43].

From the macroscopic model, the fuel cell performance has been dramatically improved by the optimised design of the GDL. The benefits of the new GDL design translate into performance gains of up to approximately 80% for the macroscopic PEMFC due to the significant improvements in EEC and ETC. In addition, the increase in ETC leads to a reduction in membrane temperature, leading to a higher membrane water content. The result is an increase in membrane proton conductivity, which increasingly reduces the ohmic overpotential and dramatically improves the fuel cell performance.

#### **3.4 Hydrogen storage**

As the bridge between the production and utilisation of hydrogen, hydrogen storage technology runs through the hydrogen end of the industry chain to the fuel cell end. It is an important link in controlling the cost of hydrogen.

How hydrogen is stored is of great concern. Hydrogen has a high energy density, three times that of petrol; it is light, weighing only 1 kg for 11*:*2*m*3; it is very easy to dissipate because it is far less dense than air; and since hydrogen is a liquid with a density of 70*:*78*kg=m*<sup>3</sup> at 253°C, it is nearly 850 times denser than hydrogen under standard conditions (approx. 0*:*08342*kg=m*3). Therefore, cryogenic liquid hydrogen storage is a highly desirable form of hydrogen storage in terms of energy storage density alone. However, there are still some problems with low-temperature liquid hydrogen storage technology. Firstly, the process of hydrogen liquefaction consumes much energy, with the actual energy consumption being equivalent to 30% of the total hydrogen energy; secondly, liquid hydrogen storage tanks require high selection criteria for insulation materials and tank design due to issues such as sealing, insulation and safety, leading to increased manufacturing difficulties and high costs.

The abovementioned problems limit the use of hydrogen as a fuel in aviation. A new hydrogen storage tank concept has been proposed to solve this problem, using cryogenic liquid nitrogen as an insulating layer. The structure is shown in **Figure 16**. This is a multi-layer isolation structure with a chamber between the tank walls that can be filled with cryogenic liquid nitrogen, which is used to maintain the temperature of the liquid hydrogen inside the tank. The liquid nitrogen can be replenished simultaneously as the fuel is refuelled.

To better analyse this concept and explore its actual performance, a modelling analysis of the structure of the hydrogen storage tank was carried out. It is first assumed that this tank is designed to meet the following mission requirements. A regional aircraft capable of carrying 32 passengers, with a required range of 2100 km, an altitude of 9144 m, a cruise speed of Mach 0.65 and a fuel load (liquid hydrogen) of 1150 kg. To compare with the original liquid hydrogen storage equipment, two more advanced types of hydrogen storage tanks (Polyurethane (PU) foam and vacuumbased multilayer insulation (MLI)) were also analysed.

The analysis of the thermodynamic properties, insulation capacity and adaptability to different environmental conditions of the three tanks leads to the conclusion that

*Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

**Figure 15.** *Effects of the fibre angle on (a) the EEC, (b) the ETC and (c) the REEC and RETC.*

**Figure 16.** *Schematic description of the new concept of the hydrogen storage tank.*

the novel concept hydrogen storage tank can significantly limit the tank's volume by introducing liquid nitrogen to limit the heat flux into the tank. The volume savings compared with foam materials increase with the reduction of the liquid hydrogen payload. For small tanks, the volume savings compared with pure foam insulation, as well as the increased adaptability and robustness, come at the cost of the tank weight, as additional structural walls are necessary. MLI is the best performer of all concepts in tank volume because it has the lowest insulation thickness. For larger tanks, however, the MLI has a comparative advantage over the novel concept in weight reduction. In the 12-hour overnight stop design, the MLI outperforms the other two concepts. But the novel concept still fulfils the challenge of the 12-hour stop. The novel concept is expected to be a reliable hydrogen storage system in the future because it does not need to maintain vacuum conditions and its reliability in the face of active thermal system failure.

## **3.5 Reliability**

Safety is the most critical factor for a civil airliner; therefore, the safety requirements of such aeroplanes are extremely stringent according to national regulations. However, higher safety redundancy means lower specific power for the fuel cell. The increase in specific power for large civil aircraft fuel cell systems is sought from different approaches rather than focusing on the design of the fuel cell system. Having quantified the extent to which aviation safety specifications affect the specific power of fuel cell systems, strategies to reduce these effects need to be investigated.

The safety certification guidelines for aircraft guide the design of fuel cell systems. To meet the requirements of safety certification guidelines, it is necessary to design the fuel cell system with sufficient redundancy. However, the more redundancy there is in a system, the heavier it becomes, thus reducing the specific power of the fuel cell system and payload. The theoretical redundancy (*DORtheo*) can be calculated using the following formula.

$$DOR\_{theo} = \log\_{(1-W)}(p) \tag{4}$$

As the DOR follows a logarithmic function, a considerable increase in reliability is required to reduce the DOR significantly. The relationship between redundancy and failure rate is given in **Figure 17**. It can be seen that the sensitivity of DOR and the sensitivity of different powers to system reliability increase as system reliability increases.

Subsystem redundancy analysis of fuel cells helps reduce the overall system's mass, as less redundancy is required for reliable components than for the system as a whole. Further improvements in fault-tolerant systems are a good direction, such as the new PEMFC stacks, which can bypass faulty cells [44].

## *3.5.1 Time constant determined extended operations (TCD-ETOPS)*

TCD-ETOPS is a newly introduced concept based on ETOPS (Extended Operations), which allows the aircraft to maintain a greater distance from diversionary airports, allowing shorter flight paths and operations over water and in remote areas [45]. And the purpose of TEC-ETOPS is to reduce the impact of safety certification on

**Figure 17.**

*Degree of redundancy as a function of failure rate. Area colours show different failure modes. Green: minor, blue: major, red: hazardous, white: catastrophic.*

#### **Figure 18.**

*Visualisation of the time constant determined extended operations (TCD-ETOPS). The failure condition is met at 10% initial power loss.*

the specific power of the system. The time constant for a given degradation mechanism represents the time scale of power loss of the FC. If the diversionary airport is within the TEC-ETOPS range, the aircraft can continue flying and landing in the event of a subsystem failure. Suppose the time scale of degradation allows the aircraft to reach the destination airport without further deterioration of the failure. In this case, a greater incidence of failure can be allowed with a guarantee of safety; therefore, the system's redundancy can be reduced.

For example, as shown in **Figure 18**, the range of the aircraft's TCD-ETOPS after a failure encompasses the time the plane arrives at the destination airport. This failure has a minimal impact on the aircraft's operations, and then there is a possibility that the failure of this subsystem can be degraded to a lower failure level.

The PEMFC system can fail to power an aircraft due to several failure conditions, the 10 most serious being: overheating, leakage, fracture, extruded, stress corrosion cracking, erosion, deposition, cavitation, inadequate structural support and failure to perform its function [46]. But some of the degradations that occur are also reversible; for example, the loss of voltage due to low humidity in the ground is reversible to some extent [47], while the supply of pure hydrogen to the anode after carbon monoxide poisoning of the membrane restores the total voltage [48]. In this way, the failed FCs are restored after TCD-ETOPS, then repair or replacement of the FCs can be postponed or avoided, and if in flight, the range of TCD-ETOPS in flight can also be extended.

By proposing the TDC- ETOPS concept, the impact on the specific power of the fuel cell due to excessive redundancy requirements can be compensated. By analysing the fuel cell system at the component level, there is also the prospect of reducing the system's mass. In future research, it is also necessary to address the different failure modes of PEMFCs and the degradation relationships caused by failures, thus being a more detailed fuel cell system design.

## **4. Summary and outlook**

As early as the 1960s, fuel cells were used in the NASA Gemini spacecraft, followed by the launch of fuel cell concept models by many internationally renowned car companies. In the twenty-first century, as the development of hydrogen energy

## *Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

technology has gradually matured, the world's major developed countries have attached great importance to the development of the hydrogen energy industry, and hydrogen energy has become an important strategic choice to accelerate energy transformation and upgrade and cultivate new economic growth points.

Fuel cells have the following advantages.


Although the fuel cell presents many attractive advantages, it also has severe shortcomings. The main bottleneck for fuel cell applications is the high cost. Due to cost constraints, fuel cell technology is currently only economically competitive in a few specific applications (e.g. on space vehicles).

Power density is another significant limitation. Power density represents the power produced per unit volume (volumetric power density) or per unit mass (mass power density) of a fuel cell.

Although the power density of fuel cells has increased significantly over the last few decades, it needs to be increased further if they are to be competitive in portable electronics and the automotive sector. Internal combustion engines and ordinary batteries often outperform fuel cells in terms of volumetric power density, while they are very close in terms of mass power density.

The availability and storage of fuel pose an even more profound challenge. Fuel cells work best when fuelled by hydrogen, but hydrogen is not readily available, has a low bulk energy density and is challenging to store. Other alternative fuels are difficult to use directly and often require reforming. All of these issues reduce the performance of the fuel cell and increase the requirement for auxiliary equipment. Thus, although gasoline is an attractive fuel from an energy density point of view, it is not suitable for fuel cell use.

For the time being, the hydrogen energy sector is still in its early stages of development and is an integral part of the zero-carbon, low-carbon era. But hydrogen has strengths and weaknesses in equal measure, so it needs to be used for its strengths and avoided for its shortcomings.

## **Author details**

Xin Gao1,2\* and Chengwei Zhao2

1 "Fuel Cells for Aviation" of Excellence Cluster SE2A, Technische Universität Braunschweig, Braunschweig, Germany

2 Institut für Energie- und Systemverfahrenstechnik (InES), Institute of Energy and Process Systems Engineering, Technische Universität Braunschweig, Braunschweig, Germany

\*Address all correspondence to: xin.gao@tu-braunschweig.de

© 2023 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

*Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

## **References**

[1] Sharaf OZ, Orhan MF. An overview of fuel cell technology: Fundamentals and applications. Renewable and Sustainable Energy Reviews. 2014;**32**: 810-853

[2] European Commission, Directorate-General for Mobility and Transport, Directorate-General for Research and Innovation, Flightpath 2050: Europe's Vision for Aviation: Maintaining Global Leadership and Serving Society's Needs. Publications Office; 2011. DOI: 10.2777/ 50266

[3] CO2 emissions from Commercial Aviation. International Council on ... (no date). 2018. Available from: https://the icct.org/sites/default/files/publications/ ICCT\_CO2-commercl-aviation-2018\_ 20190918.pdf. [Accessed: November 16, 2022]

[4] Noland JK. Hydrogen electric airplanes: A disruptive technological path to clean up the aviation sector. IEEE Electrical Insulation Magazine. 2021; **9**(1):92-e102. DOI: 10.1109/ MELE.2020.3047173

[5] Castro AL, Lacava PT, Mourao CHB. Feasibility of using fuel cells in a small aircraft. In: AIAA Aviation 2021 Forum. Reston, Virginia: American Institute of Aeronautics and Astronautics; 2021

[6] Wang B et al. Current technologies and challenges of applying fuel cell hybrid propulsion systems in unmanned aerial vehicles. Progress in Aerospace Sciences. 2020;**116**:100620. DOI: 10.1016/j.paerosci.2020.100620

[7] Huang Y, Xiao X, Kang H, Lv J, Zeng R, Shen J. Thermal management of polymer electrolyte membrane fuel cells: A critical review of heat transfer

mechanisms, cooling approaches, and advanced cooling techniques analysis. Energy Conversion and Management. 2022;**254**:115221. DOI: 10.1016/ j.enconman.2022.115221

[8] Parent KE. Fuel Cells: Energy from Gases Instead of Gasoline [Internet]. acs. org. Available from: https://www.acs. org/content/dam/acsorg/greenche mistry/education/resources/fuel-cells-e nergy-from-gases-instead-of-gasoline. pdf. [Accessed: Nov 2022]

[9] Kurzweil P. Renaissance Der Brennstoffzelletechnik. In: Brennstoffzellentechnik Grundlagen, Komponenten, Systeme, Anwendungen. Wiesbaden, Germany: Springer Vieweg; 2013. pp. 8-9. DOI: 10.1007/978-3-658- 00085-1

[10] Wee J-H. Which type of fuel cell is more competitive for portable application: Direct methanol fuel cells or direct borohydride fuel cells? Journal of Power Sources. 2006;**161**(1):1-10. DOI: 10.1016/j.jpowsour.2006.07.032

[11] https://en.wikipedia.org/ wiki/Proton-exchange\_membrane\_fue l\_cell#History

[12] Rathke P, Kallo J, Schirmer J, Stephan T, Waiblinger W, Weiss-Ungethu ̈m J. Antares DLR-H2 e flying test bed for aircraft fuel cell systems development. ECS Transactions. 2013; **51**(1):229-e41. DOI: 10.1149/05101. 0229ecst

[13] DLR - zero-emission air transport [Internet]. DLRARTICLE DLR Portal. Available from: https://www.dlr.de/ content/en/articles/news/2016/ 20160929\_zero-emission-air-transportfirst-flight-of-four-seat-passenger-airc

raft-hy4\_19469.html. [Accessed: Nov 2022]

[14] Hydrogeit. Hy4 gets permit to fly [Internet]. H2. 2021. Available from: https://h2-international.com/2021/02/ 12/hy4-gets-permit-to-fly/. [Accessed: Oct 2022]

[15] Kandlikar S, Garofalo M, Lu Z. Water management in a pemfc: Water transport mechanism and material degradation in gas diffusion layers. Fuel Cells. 2011;**11**:814-823

[16] Chen J, Matsuura T, Hori M. Novel gas diffusion layer with water management function for pemfc. Journal of Power Sources. 2004;**131**:155-161

[17] Wang B, Deng H, Jiao K. Purge strategy optimisation of proton exchange membrane fuel cell with anode recirculation. Applied Energy. 2018;**225**: 1-13

[18] Halseid R, Vie PJ, Tunold R. Effect of ammonia on the performance of polymer electrolyte membrane fuel cells. Journal of Power Sources. 2006;**154**:343-350

[19] Imamura D, Ebata D, Hshimasa Y, Akai M, et al. Impact of Hydrogen Fuel Impurities on PEMFC Performance. SAE Technical Paper 2007-01-2010. 2007. DOI: 10.4271/2007-01-2010

[20] Wang H, Turner J. Reviewing metallic pemfc bipolar plates. Fuel Cells. 2010;**10**:510-519

[21] Haider R et al. High temperature proton exchange membrane fuel cells: Progress in advanced materials and key technologies. Chemical Society Reviews. 2021;**50**(2):1138-1187. DOI: 10.1039/ d0cs00296h

[22] Araya SS, Zhou F, Liso V, Sahlin SL, Vang JR, Thomas S, et al. A

comprehensive review of pbi-based high temperature pem fuel cells. International Journal of Hydrogen Energy. 2016;**41**: 21310-21344. [Accessed: Nov 2022]

[23] Hartnig C, Schmidt TJ. On a new degradation mode for high-temperature polymer electrolyte fuel cells: How bipolar plate degradation affects cell performance. Electrochimica Acta. 2011; **56**:4237-4242

[24] Deutsches Zentrum für Luft- und Raumfahrt. Emissionsfreier Antrieb für die Luftfahrt: Erstflug des viersitzigen Passagier- flugzeugs HY4. 2016. Available from: https://www.dlr.de/c ontent/de/artikel/news/2016/ 20160929\_emissionsfreier-antrieb-fuerdie-luftfahrt-erstf lug-des-viersitzigenpassagierflugzeugs-hy4\_19469#/gallery/ 24480. [Accessed: November 26, 2020

[25] Srinath AN, Pena Lopez A, Miran Fashandi SA, Lechat S, Di Legge G, Nabavi SA, et al. Thermal management system architecture for hydrogenpowered propulsion technologies: Practices, thematic clusters, system architectures, future challenges, and opportunities. Energies. 2022;**15**(1):304. DOI: 10.3390/en15010304

[26] Kadyk T, Schenkendorf R, Hawner S, Yildiz B, Ro€mer U. Design of fuel cell systems for aviation: Representative mission profiles and sensitivity analyses. Front Energy Res. 2019;**7**:35. DOI: 10.3389/ fenrg.2019.00035

[27] NASA. U.S. Standard Atmosphere. Available from: https://ntrs.nasa.gov/a pi/citations/19770009539/downloads/ 19770009539.pdf. [Accessed: May 31, 2022]

[28] Kozulovic D. Heat release of fuel cell powered aircraft. In: Proceedings of Global Power and Propulsion Society. 2020. DOI: 10.33737/gpps20-tc-99

*Clean and Sustainable Hydrogen-Electric Propulsion DOI: http://dx.doi.org/10.5772/intechopen.109215*

[29] Piancastelli L, Frizziero L, Donnici G. The Meredith ramjet: An efficient way to recover the heat wasted in piston engine cooling. Journal of Engineering and Applied Science. 2015; **10**(12):5327-5333

[30] Martinez-Val R, Perez E, Palacin J. Historical perspective of air transport productivity and efficiency. In: 43rd AIAA Aerospace Sciences Meeting and Exhibit. 2005. DOI: 10.2514/6.2005-121

[31] Chen L, Kang Q, Tao W. Pore-scale study of reactive transport processes in catalyst layer agglomerates of proton exchange membrane fuel cells. Electrochimica Acta. 2019;**306**:454e65. DOI: 10.1016/j.electacta.2019.03.158

[32] Ryan EM, Mukherjee PP. Mesoscale modelling in electrochemical devicesda critical perspective. Progress in Energy and Combustion Science. 2019;**71**:118 e42. DOI: 10.1016/j.pecs.2018.11.002

[33] Chen Q, Niu Z, Li H, Jiao K, Wang Y. Recent progress of gas diffusion layer in proton exchange membrane fuel cell: Two- phase flow and material properties. International Journal of Hydrogen Energy. 2021;**46**:8640-8e71. DOI: 10.1016/j.ijhydene.2020.12.076

[34] Sui PC, Djilali N. Analysis of coupled electron and mass transport in the gas diffusion layer of a PEM fuel cell. Journal of Power Sources. 2006;**161**:294-e300. DOI: 10.1016/j.jpowsour.2006.03.079

[35] Bazylak A. Liquid water visualisation in PEM fuel cells: A review. International Journal of Hydrogen Energy. 2009;**34**: 3845e57. DOI: 10.1016/j.ijhydene. 2009.02.084

[36] Um S, Wang C-Y, Chen KS. Computational fluid dynamics modelling of proton exchange membrane fuel cells. Journal of the Electrochemical Society. 2000;**147**:4485. DOI: 10.1149/1.1394090

[37] Shen J, Tu Z, Chan SH. Performance enhancement in a proton exchange membrane fuel cell with a novel 3D flow field. Applied Thermal Engineering. 2020;**164**:114464. DOI: 10.1016/j. applthermaleng.2019.114464

[38] Kulikovsky AA, Divisek J, Kornyshev AA. Modeling the cathode compartment of polymer electrolyte fuel cells: Dead and active reaction zones. Journal of the Electrochemical Society. 1999;**146**:3981-3e91. DOI: 10.1149/ 1.1392580

[39] Wu W, Jiang F. Microstructure reconstruction and characterization of PEMFC electrodes. International Journal of Hydrogen Energy. 2014;**39**: 15894-1e906. DOI: 10.1016/J. IJHYDENE.2014.03.074

[40] Nabovati A, Hinebaugh J, Bazylak A, Amon CH. Effect of porosity heterogeneity on the permeability and tortuosity of gas diffusion layers in polymer electrolyte membrane fuel cells. Journal of Power Sources. 2014;**248**:83 e90. DOI: 10.1016/J.JPOWSOUR. 2013.09.061

[41] Zamel N, Li X, Shen J. Numerical estimation of the effective electrical conductivity in carbon paper diffusion media. Applied Energy. 2012;**93**:39-e44. DOI: 10.1016/J.APENERGY.2011.08.037

[42] Yablecki J, Bazylak A. Determining the effective thermal conductivity of compressed PEMFC GDLs through thermal resistance modelling. Journal of Power Sources. 2012;**217**:470-4e8. DOI: 10.1016/j.jpowsour.2012.06.011

[43] Lange KJ, Sui PC, Djilali N. Pore scale modeling of a proton exchange membrane fuel cell catalyst layer: Effects of water vapor and temperature. Journal of Power Sources. 2011;**196**:3195-e203. DOI: 10.1016/j.jpowsour.2010.11.118

[44] Scott P, Chen Y, Calay R, Bhinder F. Experimental investigation into a novel modular PEMFC fuel cell stack. Fuel Cells. 2015;**15**:306-321. DOI: 10.1002/ fuce.201200212

[45] Airbus: Getting to Grips with ETOPS: Flight Operations & Line Assistance. 1998. Available from: https:// www.737ng.co.uk/AIRBUS%20ETOPS% 20Guide.pdf. [Accessed: November 26, 2020]

[46] Bahrebar S, Blaabjerg F, Wang H, Vafamand N, Khoo-Ban M-H, Rastayesh S, et al. A novel type-2 fuzzy logic for improved risk analysis of proton exchange membrane fuel cells in marine power systems application. Energies. 2018;**11**:721. DOI: 10.3390/en11040721

[47] Knights SD, Colbow KM, St-Pierre J, Wilkinson DP. Aging mechanisms and lifetime of PEFC and DMFC. Journal of Power Sources. 2004;**127**:127-134. DOI: 10.1016/j.jpowsour.2003.09.033

[48] Le Canut J-M, Abouatallah RM, Harrington DA. Detection of membrane drying, fuel cell flooding, and anode catalyst poisoning on pemfc stacks by electrochemical impedance spectroscopy. Journal of Power Sources. 2006;**153**:A857. DOI: 10.1149/1.2179200

## **Chapter 5**

## Cyclic Oxidation of Diffusion Aluminide Coatings

*Marta Kianicová*

## **Abstract**

The diffusion aluminide coatings are used for high-temperature applications. Structural materials of particular components degrade during service due to fatigue, creep, oxidation, corrosion and erosion. The requirements of higher efficiency of modern industrial applications increase the development of new structural materials, technologies and protective coatings. Properties of many structural materials such ultimate tensile strength, creep strength and fatigue are generally optimized for maximum high-carrying loading with less emphasis on environmental resistance. For these applications, the performance characteristics are limited by the operating conditions, which can be tolerated by the used materials. The main structural materials for high mechanical and thermal loading are superalloys protected against aggressive environment by coatings. Cyclic oxidation is the superposition of thermal cycles in an oxidation environment. The main goal of the experimental work was to compare the cyclic oxidation of protective Al and AlSi coatings deposited on both Inconel 713 LC and MAR-M247 superalloys. The resulting graph revealed that samples from IN 713 LC without coating show good resistance and their mass change is maintained above zero limit. Samples from MAR 247 LC with both Al and AlSi coatings appear to be the most acceptable selection of combination relating to superalloys/coating.

**Keywords:** cyclic oxidation, superalloys, diffusion aluminide coating, CVD deposition

## **1. Introduction**

Increasing the technical level, utility values, quality and reliability of engineering components belong to the main directions of economic development. The surface treatment also plays a significant role in the final quality of the structural unit. An inappropriate choice of surface treatment, non-compliance with the technological procedure or just an unaesthetic appearance can devalue an otherwise excellent technical work.

Surface engineering includes those processes that make it possible to modify the surfaces of structural components with the aim of improving primarily their mechanical and utility properties and, last but not least, their esthetic appearance, or to reduce the economic costs of production. It is therefore a modification of the surface of the component, which does not affect the bulk properties. The main goal is to increase the service life of components and the operational reliability of equipment. It is often aimed at increasing corrosion resistance, surface sliding wear resistance, providing a diffusion barrier, thermal insulation, etc. while reducing production costs, especially energy savings with an impact on minimizing environmental pollution.

There are several ways to treat metal surfaces, which can be divided into three main categories:


## **1.1 Modifying the surface without altering chemical constitution.**

A. Heat treatment

Heat treatment is the practical application of physical metallurgy with the aim of modifying the material structure to obtain desired properties. Strengthening by phase transformation essentially belongs to thermal technology processing. Surface hardening is a common method of changing the mechanical properties of surface layers especially for steels. The purpose of this treatment is to increase hardness and wear resistance component surface. The conditions of transformation or breakdown of austenite are decisive depending on temperature, time and rate of cooling. To austenitize the surface layer, it is necessary to perform heating at a higher rate than the heat is dissipated into the part. Phase diagrams are used to understand the heat treatment process. For example, an Isothermal diagram (IT) describes the formation of austenite, Time Temperature Transformation (TTT) diagram.describes the decomposition of austenite, Continuous Cooling and Continuous Heating diagrams are commonly referenced to as CCT and CHT diagrams, which are also used to understand the effects of heating and cooling of steel during heat-treatment process [1]. In practice, heating is commonly used by flame or electrical induction. Recently, they have been focusing on themselves for changes induced by laser irradiation or electron

beam techniques [2–6]. Surface heat treatments by laser are manufacturing technologies that are gaining industrial interests inthe last years.

Laser radiation makes it possible to concentrate the power density up to 10<sup>16</sup> W.m<sup>2</sup> and deliver considerable amount of energy to the treated surface contactlessly and quickly. To the characteristic energy parameters of radiation, pulse length or speed of movement have a great influence on the temperature field temperature source (for a continuous laser), temperaturephysical characteristics and geometric dimensions processed body. Depending on the power density and duration of action, either only heating and cooling in the solid phase takes place in the surface layer, or there is melting of the surface and the possibility of forming an amorphous state during particularly rapid cooling. Rapid temperature and structural changes cause thermal stresses, which can also result in the formation of cracks. Lasers are particularly suitable for curing relatively small or inaccessible surfaces. Laser is considered a useful technique since it offers control over case depth, uniform microstructure, uniform hardness, and minimum material distortion. Laser transformation hardening of steels is a diffusion-less transformation process and eliminate the formation of pearlite phases [7].

In addition, the interestingness of this process lies in the possibility of direct integration of a very flexible laser heat source without the use of a quenching medium, as well as in the possibility of producing various microstructures with a hard surface and residual compressive stresses. Electron beam techniques have lower investment and operating costs.

## B. Mechanical treatment

**Cold working** the surface by shot peening or other specialized surface treatments were designed to enhance the surface integrity and service life of structural components. Most of these mechanisms are very sensitive to microstructure. Surface treatments like Ultra Sonic Peening, Laser Shock Peening, and Shot Peening are used to enhance the surface microstructural properties that help in improving the service life. These applications lead to deformed surface layers and increase the stored energy, form compressive residual stresses due to higher density of defects in the crystal structure, increase the hardness, fatigue life and stress corrosion resistance. Residual compressive stresses and increased surface hardness are the key outcomes of shot peening [8–10]. The simulation methodology and numerical prediction are used to optimize microstructure and shot peening parameters to reduce the relaxation of compressive residual stresses during service life [11].

## C. Thermomechanical treatment

This process consists of a combination of heat treatment and plastic deformation. Thermomechanical processing is based on a combination of several operations such as deformation, heating and cooling performed in different cycles, which further determine their classification. A high density of structural imperfections such as dislocations, vacancies or layer defects and their

distribution due to deformation mechanisms affect the processes of structural changes and thus the resulting properties and integrity of the surface of the components. However, structural changes generate new ones and redistribute the original ones. Thus, the mechanism and kinetics of these structural changes depend on the nature and density of imperfections and consequently affect the number and distribution of such imperfections. Thermomechanical treatment is although used to improve the mechanical properties of alloys by introducing twins and nano-precipitation phase in the lamellar structure of the homogenized alloys [12]. Multi-stage thermomechanical processing with the stepwise temperature reduction can improve not only the mechanical but also the electrical properties of the alloy [13].

## **1.2 Altering the chemistry of surface regions of the substrate**


Nowadays, the structural materials made for marine, aerospace engineering, transport vehicles must be protected against corrosion, abrasion, or high temperature oxidation and dense and adhesive oxide coatings improve the resistant of substrates against these degradation modes and enhance the surface chemical and mechanical properties [22]. Nanocoatings developed in thin layer having scale 1–100 nanometers thickness offer much better processing properties than conventional coatings. They provide effective solution against chemical or mechanical effects. Nanocoatings are not protective layers but their particles bind themselves both physically and chemically to the substrate surface and give it very good protection. Nanocoatings have huge application in different industries except for automobile, marine or aircraft industries, also in defense, electronic or medical industries [23, 24].


impacted into a solid. Ions impinging on the component cause many chemical and physical changes of materials transferring energy and momentum to electrons causing a structural change. The ion energy and the target composition determine the depth of ion penetration. Some studies reported that some of ions have antimicrobial effect and its nanoparticles act as antimicrobials [28].

## **1.3 Surface treatment by coating deposition**

The current age is the age of surface treatment technologies that allow countless possibilities of obtaining coatings with multiple properties. State-of-the-art manufacturing techniques are used in diverse areas, such as optical memories and filters, fiber lasers, LED displays, various implants or environmental devices.


structures. It has been shown that galvanized bar slows down cracking during service and concrete spalling is less likely or delayed [34].


The preceding brief review shows an extraordinary extend and variability of methods utilized in the surface engineering. Hereafter, we will focuse on some important kinds of coatings on metallic surfaces and the processes of their degradation in service. Metals as the primary construction materials used in industry are exposed to oxidation and to active chemical environment. To ensure the quality and protection of metal products, different kinds of metal coatings are employed in the industry. There are two basic determinants of metal coatings.

**The first class** of coatings serves as a protection against the aggressive environment, i.e., they increase the resistance to thermal oxidation, corrosion and erosion.

**The second class** represents layers obtaining by surface hardening which increases surface strength for load carrying, surface hardness of a substrate material and introduce compressive residual stresses.

Thus, with respect to their function, the coatings can be divided into two main classes:


Diffusion aluminide coatings (DACs), overlay coatings and thermal barrier coatings (TBCs) are typical representatives of the first class. They improve the resistance to high temperature creep, fatigue, thermal oxidation, corrosion and erosion. Nickel and cobalt based superalloys are mostly utilized as substrate materials for turbine blades, blade rings, stator/rotor discs, etc. From the point of view of corrosive and oxidizing effects, to the monolayer protective coatings belong DAC and overlay coatings. DAC are based on the intermetallic compound β-NiAl that forms under the influence of the substrate (usually Ni superalloys). The technological procedures lie in various methods of deposition of an equally balanced suspension of aluminum- and silicon powders on the substrate surface together with activators (e.g., NH4Cl) and organic binders (e.g. coloxylin). Diffusive tempering is applied at about 1000°C in an inert argon atmosphere or vacuum with the dwell time of several hours and subsequent slow cooling in the retort. As shown in **Figure 1**, the resulting microstructure of the layer consists of two sublayers – the outer layer (OL) and the diffusion zone (DZ). The microstructure of OL is composed of the β-NiAl phase with small number of complex phases and carbides Al-Cr-Ni, Mo-Cr-Nb. The DZ is formed by β-NiAl with

**Figure 1.** *Corrosion and oxidation resistance of metallic coatings on superalloys.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

many Cr-Mo, Mo-Cr-Nb rich particles. DACs are an integral part of the substrate material with just a slightly higher hardness. They primarily role is to protect the surface from high-temperature oxidation, corrosion and erosion but they should not reduce the resistance to creep, cyclic creep and thermo-mechanical fatigue [37]. On the other hand, the composition of the overlay coatings (known as MCrAlX) remains independent of the substrate alloy.

The most popular representatives of the second class are carburized or nitrided surface layers (CNSLs) produced by gas diffusion or plasma deposition processes - the case hardening. This process increases the surface hardness of a substrate material and introduces compressive residual stresses which results in a higher resistance to fatigue and wear of low-carbon steels, the materials mostly used in various dynamically loaded components as gears and shafts at room temperature. All the above mentioned components are subjected to cyclic loading in bending (tension) or, most frequently, to a combined bending (tension)/torsion loading.

## **2. Diffusion aluminide coatings**

The diffusion aluminide coatings belong to the wide category of coatings used for high-temperature applications. Structural materials of particular components degrade during service due to fatigue, creep, oxidation, corrosion and erosion. With requirements of higher efficiency of modern industrial applications increase the development of new structural materials, technologies and protective coatings. Properties of many structural materials such ultimate tensile strength, creep strength and fatigue are generally optimized for maximum high-carrying loading with less emphasis on environmental resistance. For these applications, the performance characteristics are limited by the operating conditions which can be tolerated by the used materials. The main structural materials for high mechanical and thermal loading are nickel and cobalt-base superalloys protected against aggressive environment by coatings [38–39]. The coating systems currently in use particularly in gas turbine blade applications can be divided into three generic groups [40]:


The coatings deposited on rotor turbine blades provide an optimal protect in the range of specified lifetime against destructive effects of high-temperature corrosion, oxidation and erosion provided the following requirements are satisfied [41]:

• High oxidation and corrosion resistance.

The coating must be thermodynamically stable and creates protective thin oxide film of uniform thickness. Slow growth rate of protective and adherent scale is desirable. In accordance with these requirements the coating should contain higher content of element (Al, Cr) to able to form protective scales.


Diffusion aluminide coatings have been designed to protect nickel-base superalloys at elevated temperatures and they are based on simple aluminide or modified aluminide compounds. DACs and modified DACs involve the surface layers of substrate alloy which is enriched with aluminum, chromium, silicon, platinum, zirconium, palladium, yttrium or hafnium to form protected oxides through diffusion and/ or enhance TGO adherence [42]. Their thickness is in the range of (10–100) μm. These elements react with the main constituents of substrate and form intermetallic compounds (aluminides) with significant part of element comprising protected oxide. For nickel-base superalloys NiAl (**Figure 2**) is the main aluminide compound of coating which produces corrosion and oxidation resistant thermally grown oxide.

**Figure 2.** *Binary phase diagram Al-Ni.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

The selection of the appropriate coating composition depends on requirements of the corrosion/oxidation resistance and the environment. The main alloying elements in metallic coatings are briefly described below [43, 44]:

**Nickel** is the base element in DACs in Ni base substrates because it depresses chemical activity of Al and minimizes the interdiffusion between coating and substrate.

Aluminum form on the top of coating oxide Al2O3 which protects against oxidation up to 1200°C.

**Chromium** reduces critical level of Al needed to form alumina and chromium oxides protect coating against hot corrosion and oxidation up to 900°C.

**Silicon** is effective against low temperature hot corrosion and promotes formation of Al2O3.

**Platinum** increases TGO adherence as well as reduces oxidation kinetics. The presence of Pt in a solid solution, or as two-phase coating, provide resistance to I. type of hot corrosion; if a continuous PtAl2 phase is present, coating is resisting against II. type of hot corrosion.

**Zirconium** has beneficial effect on the adhesion of TGO, it hinders the outward diffusion of Al cations.

**Palladium** provides excellent high temperature oxidation resistance and scale adhesion by reducing or preventing void growth, mitigating the detrimental effects of sulfur and accelerating alumina scale healing after spallation. Pd replaces Pt in aluminide coatings due to high costs of Pt.

**Yttrium** is a typical oxygen-active element which has beneficial effect on corrosion behavior of coatings.

**Hafnium** additions to aluminides markedly improve the adhesion of thermally grown oxides thereby improving the high-temperature oxidation performance.

DACs are the most spread types of coatings for turbines blades. Their disadvantage is strong dependence on substrate composition. For the application in aerospace industry is very important to find the critical ductile-to-brittle transition temperature (DBTT) which limits the utilization. Material undergoes under DBTT to failure by brittle fracture. The DBTT of diffusion aluminide coatings on superalloys varies between (700 up to 900) °C [45]. Some of the factors influencing the DBTT are coating process, composition, microstructure and phase distribution. With decreasing content of Al, thickness and roughness of coating DBTT decreases. DBTT of NiAl phase is reduced by more than 100°C when Al content is lowered from 32 to 25 wt. %.

## **2.1 Technologies of deposition**

Results of many examinations have confirmed that aluminide coatings are characterized with very good corrosion and oxidation resistance at high-temperatures. They have homogeneous microstructure with good thermomechanical properties. In order to protect critical internal surfaces from oxidation and hot corrosion, the state of the art involves the use off diffusion coatings, formed by pack cementation, chemical vapor deposition (CVD), above-the-pack diffusion processes and/or by slurry cementation.

#### *2.1.1 Pack cementation*

Halide-activated pack cementation is a CVD process in which "pack" is made up of a mixture of a pure metal or alloy source, a halide salt activator and an inert filler material, usually Al2O3.

A filler allows the active pack constituents (the master alloy and activator salt) to be well distributed. This allows the halide vapors to easily reach the substrate surface, resulting in an overall uniform coating. The substrate to be coated is either buried inside the pack (the 'in-pack process') or suspended outside the pack (the 'out-ofpack' process). A controllable atmosphere, usually Ar or H2/Ar, surrounds the pack as it is heated at a definite process temperature ranging from 700 to 1050°C. The metallic powder reacts with the halide salt activator to form volatile metal halide species of significant partial pressures. In the case of Pt-modified aluminide coatings, components are electroplated with a thin Pt layer prior to the aluminization.

Suppose NH4Cl is used as the activator, the reaction will be given below [46]:

$$\text{M(s)} + \text{xNH}\_4\text{Cl(s,g)} = \text{MLx}(\text{g}) + \text{xNH}\_4(\text{g})\tag{1}$$

where M is the element to be coated.

For aluminizing process Al, AlCl, AlCl2 and AlCl3 vapor substances can be formed and for chromizing Cr, CrCl2, CrCl3 and CrCl4.

Partial pressure gradients support vapor transport tio the metal surface where desired coating specified phase, microstructure and composition forms via dissociation or disproportionation of the halide molecules and reaction with substrate.

On this basis, the pack cementation can be described as follows [46]:


Illustration of these events are schematically visible in the **Figure 3**. All of these events dictate the nature of the final microstructure and thickness of the surface layer.

There are several pack cementation processes, i.e. pack aluminizing, chromizing, siliconizing, or co-deposition needed for suitable surface layers. Aluminized coatings and chromized coatings are two traditional diffusion coatings that have been extensively used for almost a century.

Halide-activated pack cementation is a simple and inexpensive process useful for different structure geometries and sizes. It has great potential for using by innovative ways.

#### *2.1.2 Chemical vapor deposition*

In the CVD method two basic processes, i.e. the low-activity and high-activity process of aluminizing are distinguished. During the aluminide coating deposition process at high temperature, about 1050°C, low aluminum content NiAl phases are

#### **Figure 3.**

*Ilustration of steps occuring in the pack cementation process.*

created (low activity process), whereas at about 700°C, NiAl phases containing more aluminum are formed (high activity process).

In "low-temperature high activity" (LTHA) process the coating grows by the inward diffusion of Al and Ni2Al3 (δ-phase) forms at the matrix surface. Al further diffuses inward and the initial surface of the substrate remains at the outermost surface upon aluminizing. Alloying elements that diffuse more slowly and carbides in the substrate remain in δ-phase. Ni2Al3 is a brittle phase and furher vacuum annealing is necessary to transform it into desired thermodynamically stable NiAl (β-phase). A disadvantage of coatings obtained by LTHA process an overall thickness of structures, in some cases is not possible to use LTHA technology.

In "high-temperature low activity" (HTLA) process, thermodynamic activity of Al is lower and the coating grows by the outward diffusion of Ni, which leads to direct formation of additive β-NiAl phase above the initial surface. Interdiffusion zone forms in the substrate due to depletion of Ni and slowly diffusing elements.

Mixed coating growth mechanisms are also known and involve the diffusion of both Al and Ni [47]. This is schematically illustrated below (**Figure 4**).

An effective role of diffusion barrier between the coating and superalloy would slow the loss of Al to the substrate and limit or delay diffusion of refractory elements into the coating. Microstructure of superalloys IN 713 LC after aluminizing and heat treatment is visible in the **Figure 5**.

The CVD method gives the possibility of control of the AlXn concentration and the temperatures of the processes and through this control of aluminizing process. The HTLA CVD aluminizing process indicate superior oxidation resistance compared with the LTHA process because clean β-NiAl phase is free of precipitates and carbides.

## *2.1.3 "Out of pack" or "over-pack" method*

This method operates in a manner very similar to pack cementation; except the parts to be coated are suspended either above the pack or downstream from the pack (vapor generating) retort. The technology is based on placing the parts to be coated in a retort or vacuum furnace oven without coming into direct contact with the granulated mixture. The transfer of gases, which ensure the formation of the coating, is

#### **Figure 4.**

*Schematically illustrated microstructures in the low and high activity processes after heat treatment.*

#### **Figure 5.**

*Microstructure of coating on superalloy IN 713 LC after high activity CVD aluminizing and heat treatment.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

made possible by a neutral gas, guided by the container. Currently, several variations of the applied methods are already known (e.g. SNECMA). This process relates to the deposition of an aluminum coating on a metal part, especially on a hollow metal part comprising an internal liner [48].

The key step of the "out of pack" method is the formation of Al(Cl, F, Br)n (n < 3) halides at temperatures above 800°C which react with the surface of superalloy by the following reaction [49]:

$$Al + AlX\_3 \to AlX\_n \tag{2}$$

The AlX3 compound is called "activator". At the temperature 900-1150°C halides in the gaseous phase are reacting with substrate according to the reaction:

$$AlX\_n + Ni \to AlNi\_\mathcal{Y} + AlX\_3 \tag{3}$$

where: 1/3 < n < 3.

Activator is one of reaction products, is penetrating into the source of aluminum and is supporting the process according to the scheme:

$$Al + AlX\_3 \to AlX\_n \tag{4}$$

$$AlNi\_{\mathcal{Y}} + AlX\_3 \gets AlX\_n + Ni \tag{5}$$

All reactions are in the equilibrium and they are leading for creating the intermetallic phase AlNiy.

CVD and "out-of-pack" methods are the same from the view of chemical reactions. The difference between both methods results from the fact, that reaction: *Al* þ *AlX*<sup>3</sup> ! *AlXn* is occurring in the special generator (external) where from products are transported in the gaseous phase to the retort in which coated elements are.

#### *2.1.4 Slurry process*

Slurry aluminide coatings belong to the high temperature diffusion processes. They have been well known for a number of years and are widely used to protect metallic surfaces from oxidation and hot corrosion. Composition and microstructure of coatings formed by slurry-aluminizing process are similar to those obtained by pack cementation. Coatings made by slurry-aluminizing have high quality and the technology is characterized by smallest consuming applied materials. This allows to obtain coatings of the lowest cost and slurry method is one of the cheapest available methods of manufacture of aluminide coatings. Because of the environmental constraints of pack cementation and of out-of-pack processes related to the use of halides and other hazardous chemicals, significant efforts led to the use of water-based technologies [50, 51]. The suspension designed for deposition process can be easily modified and may be stored for a long time. In the deposition process an aluminum-containing slurry is being applied onto the component to be coated, and during heat treatment the slurry react with the substrate forming an aluminide. Slurry is composed from a powder mixture of aluminum or an aluminum alloy plus an activator along with a binder either by spraying or brushing, which is then submitted to diffusion heat treatment at temperatures usually in the 1000–1200°C range. From the point of view of the coating microstructure properties, this method can form precipitates that

## **Figure 6.** *Microstructure of Al-Si coating on nickel-based superalloy JS6K.*

inhibit the adhesion of the coating. Another problem lies on inaccessibility of some area as cooling internal passages of turbine blades. Modified aluminide coatings have been industrially developed to overcome the composition limitation that conventional slurries have. Al-Si eutectic slurries are commercially often used for elements working in hot-corrosion environments. No other components have been successfully diffused simultaneously by the slurry route. The microstructure of AlSi layer is shown in the **Figure 6** and it is composed of two sublayers. The outer Al-sublayer is created by NiAl with content of Al < 50 at. % and lower amount of phases containing refractory elements. The inner Si-sublayer creates Ni-Al matrix and higher amounts of phases of elements as Si, Cr, Mo, Ti, W and Co. The thickness of the layer varies in the range of 20–50 μm. Al-sublayer is formed by inward diffusion of aluminum what is typical for high-activity coatings. The inner Si-layer can forms during aluminizing process or operation in the high temperature environment needle-like precipitates in a β-NiAl matrix, which are oriented perpendicular to the interface substrate/diffusion zone (**Figure 7**) where is although visible coating degradation.

#### **Figure 7.**

*Microstructure of Al-Si coating on JS6K superalloy after operation in the high temperature environment.*

## **2.2 Oxidation resistance of DACs**

Protective coatings used on turbine blades were developed to serve as physical barriers between aggressive environment and the substrate. In addition, TBCs are used as thermal barriers, retard creep degradation and reduce the severity of thermal gradients. Up to now, however, no coating that would fully survive the aggressive turbine environment has been found.

The coating degrades during service at two fronts: at the coating/gas-path interface when service temperatures are below the melting point of the coating and at the coating/substrate interface at higher temperatures when diffusion mechanisms play the main role in degradation of the system substrate/coating.

The most serious degradation modes are as follows [38]:


Inter-diffusion of elements at the interface with the substrate that results in a creation of undesirable phases is, sometimes, also mentioned as an independent degradation mode.

DACs coatings have been designed to withstand three types of environmental attack: high temperature oxidation, high-temperature hot corrosion (type I) and lowtemperature hot corrosion (type II). Oxidation is a special form of corrosion degradation mode which occurs when metals and alloys are exposed to the oxygen environment. However, it should be paid regard to that the oxidation of the DACs at the coating/gas-path interface results in the formation of a protective oxide scale and in respect thereof high-temperature oxidation is not explicitly a degradation mode. If the formed oxide scale is thin, slow-growing, and adherent, it protects the substrate from further oxidation and form barrier to further oxidation. If it be to the contrary oxide scales spall and substrate is exposed to the oxygen environment and suffers from consumption of metals. "Pilling–Bedworth ratio" (PBR) is an important parameter for prediction of the oxide protection properties. It was found out that if the volume of oxide is less than the volume of metal consumed in the reaction, then it is likely that a porous oxide layer will result. This criterion is effective for most metals and alloys of practical importance and by PBR we can assume if the oxide is protective or not. Based on this theory PBR for oxidation of alloys can be expressed as [52]:

$$\text{PBR}\_{\text{alloy}} = \frac{\text{Volume of a mole of B}\_{\text{x}} \text{O}\_{\text{y}}}{\text{Volume of x moles of B in the alloy}}.\tag{6}$$

It is believed that if *PBR*>1 compressive stresses are developed in the oxide scales while tensile stresses occur when *PBR*<1*:*The larger the difference of PBR from 1, the larger stresses. Although is well known that direct relation between PBR value and the level of stresses does not hold because mechanisms for growth stresses and their relaxation are complicated. PBR calculated for aluminum gave value 1.29. However, in the practice alloys are widely used for high temperature applications and values of PBR are different from those of metals. Authors in [52] calculated PBR for Ni3Al, NiAl and NiAl3 and their results indicated that the PBR for oxidation of Ni-Al alloys are larger in regard of PBR for Al metal. PBR for Ni3Al was calculated from 1.71 to 1.88; for NiAl was in the range from 1.64 to 1.78 and PBR for NiAl3 from 1.48 to 1.57.

## *2.2.1 Cyclic oxidation*

Cyclic oxidation is the superposition of thermal cycles in an oxidation environment. Alloys used at high temperatures are subjected to the operative cycles which vary widely depending of operating conditions.

More or less rapid temperature changes in oxidizing atmosphere result in thermomechanical stresses in the oxide scales which fail due to spallation. Turbine blades of aircraft engines are a typical example for working in such conditions. Turbine blades of aircraft engines are a typical example for working in these conditions.

The main parameters that determine the operating conditions of aircraft turbine blades are the gas temperature at the turbine inlet, pressure, velocity and composition of the gas flow. The gas temperature is the most important parameter determining the specific thrust and performance of an aircraft engine. The thermal composition of the gas flow at the turbine inlet is inhomogeneous; is caused by factors that determine the dynamics of the gas flow, such as the construction of the combustion chamber, the layout of the burners, or the combustion process. The stator disks and blades of the high-pressure turbine are the hottest part of the turbine, their temperature is (200-300) °C lower than the temperature of the gases, while the difference in temperature fields is based on the height of the blades and their circumference. The temperature field in the rotor part is more homogeneous and results from the high rotational speed of the impeller.

The critical elements of an aircraft turbine are the rotor blades. They determine the maximum permissible gas temperature and the lifetime of the entire engine. They are loaded with centrifugal and dynamic forces, which cause them to be strained by tension, bending and twisting. Tensile stresses are usually higher than (200-250) MPa and are different on the convex and concave sides, being the highest near the root of the blade. High temperatures and voltages, unstable load conditions and the possibility of resonant vibrations make vanes among the most complicated parts of the engine. The highest temperature load is in the upper third of the blade, where the centrifugal force is the lowest. The temperature of the gas stream can increase by up to 500°C in just a few seconds in aircraft turbines during transients or sudden regime changes. Static stresses and high temperature of the blades lead to their deformation as a result of creep. The combination of the effects of high temperature, dynamic forces, and thermal stresses causes blade failure due to thermal fatigue.

The surface of the blades is exposed to the effects of combustion products, which cause their degradation by oxidation, corrosion and damage by solid particles carried by the gas flow. The fuel combustion process takes place at the temperature of (2000– 2200) °C. Hydrocarbon combustion is the process of its oxidation by oxygen from the air, the main products of combustion are CO2 and water vapor. The gas flow at the entrance to the turbine contains, in addition to the main ones, also secondary products of combustion, such as certain amounts of CO, H2, CH4, solid particles of carbon and sulfur compounds and other elements entering the chamber from fuel or air. The

## *Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

difference in the sample of combustion products depends on the chemical composition of the fuel and on the environment (overseas, industrial, desert).

Cyclic oxidation tests are used to monitor scale adherence and the ability of scale to successfully regenerate after repetitive scale failure. The performance of materials is generally monitored by gravimetric method although others methods have been used [53]. Experimental techniques are needed to obtain both, data on the kinetics of the oxidation reactions and characterization of the scales formed during oxidation.

The most important method used to study oxidation kinetics and oxidation rate is the gravimetric method. The principle of this method is to examine weight change due to oxidation as a function of time. Heating and cooling can be quite rapid and short cycles can be used. A simple gravimetric technique involves the exposition of the sample with known area in a furnace followed by measuring the weight change at definite intervals of time, using a sensitive balance. In this method the experiment have to be interrupted every time the weight change is measured. The sample is frequently heated and cooled what causes changes in the scale behavior such as onset of scale failure, buckling, spallation and finally mass loss. **Figure 8** represents typical mass changes during thermal cycling for three types of coatings. Non-protective behavior and negative weight change during thermal cycling represents gross scale spallation occurring from the onset of the cycling test. Protective behavior of a scale for a definite time is characteristic for materials that have a limited reserve of elements (Cr, Al or Si) to form stable oxides. After certain number of cycles scale composition changes to less protective spinels which spall rapidly and mass loss is recorded. Spallation is the loss of protective oxides at the coating/oxide interface. Strains induced by stresses during repeated thermal cycling due to thermal mismatch between oxide and metal, result in crack initiation and eventual spalling of the scales. In the case that the scale is too thin to sustain a temperature gradient, not thermal shock to the oxide. If the oxide thickens sufficiently, the strain energy stored in the oxide becomes greater than that for fracture of interface, the scale spalls [54].

**Figure 8.** *Schematic diagram of typical mass changes of coatings.*

## *2.2.2 Experimental work to study of DACs oxidation behavior*

The main goal of the experimental work was to compare the cyclic oxidation of protective aluminide coatings deposited on two types of nickel superalloys, Inconel 713 LC and MAR-M247 (**Table 1**). All samples with and without aluminide coatings were exposed to cyclic oxidation. Two type of superalloys were deposited by aluminide coating and Si modified aluminide coating using CVD out-of-pack process.

Experimental samples casted from IN 713LC and MAR 247 LC for cyclic oxidation test had cylindrical shape with dimensions of 14 5 mm. Their microstructure was consisted of the γ matrix strengthened by γ´ phase with the shape of cuboidal particles of Ni3(Al, Ti) as coherent precipitates and complex carbides. Three types of samples for two kinds of superalloys were used for cyclic oxidation tests [39]:


Aluminide coatings were applied by the "out-of-pack" method and Si modified aluminide coatings were made by the method of "pack-cementation".

Disc samples in ceramic bowl (**Figures 9** and **10**) were placed into induction furnace with the temperature 1100°C. After 23 h, the tested samples were taken from furnace and they were immediately exposed to the cooling process in the air at room


#### **Table 1.**

*Chemical composition of tested superalloy (wt. %, bal. Ni).*

**Figure 9.** *Samples of IN 713 LC with and without coatings before testing.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

**Figure 10.** *Samples of MAR 247 LC with and without coatings before testing.*

temperature for 1 h and then the mass changes were measured. This description represented one oxidation cycle. Mass changes of samples were written down after each second cycle and the photos were done.

Gravimetric weight changes of cyclically tested superalloys with coatings followed weight changes during oxide formation and the results were used to compare the resistance of superalloy/coating systems to cyclic oxidic loading in a high temperature environment as well as to choose the most suitable combination for application in practice [39].

Macroscopic views of all samples after 3rd, 10th and 20th cycles are visible in the **Figures 11**–**13**. As for uncoated superalloy MAR 247 LC test of cyclic oxidation was interrupted after 13th cycle due to continuing decrease of weight.

**Figure 11.** *Overall view on all samples after 3rd cycle of testing.*

**Figure 12.** *Overall view on all samples after 10th cycle of testing.*

#### **Figure 13.**

*Overall view on all samples after 20th cycle of testing.*

The resulting graph of samples made from IN 713 LC with and without coatings is shown in **Figure 14**. From this picture, it is clear that IN 713 LC with unmodified aluminide coating had lower life-time than without coating and it is very surprising finding. Weight changes after each oxidation cycle for MAR 247 LC samples with and without coatings are visible in the **Figure 15**. We can see that life time of MAR 247 LC without coating is very low and structural components made from this type of superalloy without coating could not be used for practice application in the oxidation environment. On the basis of results in the **Figure 15** we can see that life time of samples from MAR 247 LC superalloy with aluminide Al and AlSi coatings is practically the same.

The resulting graph of all tested samples is shown in **Figure 16**. This picture revealed that samples from IN 713 LC without coating after 24-h cycles in environment of 1100°C

**Figure 14.**

*Mass change data for IN 713 LC with and without coatings achieved from thermogravimetric analyses.*

**Figure 15.** *Mass change data for MAR 247 LC with and without coatings achieved from thermogravimetric analyses.*

show good resistance and their mass change maintain above zero limit. Samples from MAR 247 LC superalloy with both aluminide Al and AlSi coatings appear to be the most acceptable selection of combination relating to superalloys/coating.

#### *2.2.3 Microstructural research of DACs in the oxidation environment*

## *2.2.3.1 Microstructural examination of IN 713 LC without coating*

Microstructure of IN 713 LC before cyclic oxidation test (**Figure 17**) consisted of the γ-solid solution strengthened by γ´ phase with the shape of cuboidal particles of

**Figure 16.** *Weight changes after each oxidation cycle for all tested samples.*

**Figure 17.** *Microstructure of IN 713 LC without coating before testing.*

Ni3(Al, Ti) as coherent precipitates (dark particles in **Figure 18**) and complex carbides (bright particles in **Figure 17**).

After 8 cycles of cyclic oxidation testing the microstructure of IN 713 LC without protective coating degraded as we can see in **Figure 19**. It was found that γ´ particles near the surface region dissolved to the solid solution and bellow is the area of larger γ´ coarsened ones. The thickness of oxide scales and area of dissolved precipitates under the surface reached on average 5.3 μm and 32.6 μm, respectively.

From **Figures 19** and **20** is clear that degradation of uncoated samples from 8 to 18 cycles continued in the sense of formed the thicker area of oxide scale (9.9 μm) and dissolved precipitates (46.4 μm). Mechanical properties of investigated alloy mainly depend on morphology, size and volume fraction of gamma prime strengthening particles. Increasing of cycle number resulted in decreasing of the γ´ phase volume fraction what is not acceptable considering strength of the alloy.

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 18.** *γ-solid solution of IN 713 LC with γ´ phase.*

#### **Figure 19.**

*(a) Microstructure of IN 713 LC (b) Detail focused on the after 8 cycles of testing microstructure near surface region.*

#### **Figure 20.**

*(a) Microstructure of IN 713 LC (b) Microstructural appearance of after 18 cycles of testing the region under oxide surface.*

Growth of oxide scale and area of dissolved precipitates continued with increase in test cycles. After 28 cycles of cyclic oxidation was observed growth of oxide scale deeper into superalloy and the oxide thickness reached 12.9 μm. The thickness of

#### **Figure 21.**

*(a) Microstructure of IN 713 LC (b) Oxide scales after 28 cycles after 28 cycles of testing of testing.*

#### **Figure 22.**

*(a) Microstructure of IN 713 LC (b) Oxide scales under surface of after 38 cycles of testing alloy after 38 cycles of testing.*

dissolved precipitates was 73.4 μm (see **Figure 21a** and **b**). 38 test cycles caused growth of oxide scales up to 17.8 μm and area of dissolved precipitates to 100.1 μm (**Figure 22a** and **b**).

Chemical contents of the γ-solid solutions of elements in nickel (dissolved precipitates areas) in all samples after cyclic oxidation tests were roughly identical. Contents of Al were from 7.5 at. % to 8.5 at. % in the samples after 8, 18, 28 and 38 cycles; Ni ranged between 73.1 at. % and 73.4 at. %; Cr was from 15.1 at. % to 15.3 at. %. Bellow this zones were areas of coarse precipitates which contained a bit higher amount of Al (9.1 at. % – 9.8 at. %), a lower content of Cr (12.8 at. % – 13.4 at. %) and an identical content of Ni. Cross sections of oxide scales of all samples after testing (**Figure 23**) revealed two zones; inner oxide layer and outer oxide layer. It was found that the outer oxide zones were composed of lower contents of Al (about 59 at. %) compared to the inner zone (about 96 at. % Al), higher contents of Cr (about 4.9 at. %) with respect to that in the inner ones (about 0.6 at. %) and a considerably higher content of Ni in the outer layers (about 32 at. %) compared to that in the inner layers (about 2.5 at. %). From these examinations results that the degradation of uncoated samples from IN 713 LC superalloy started with a creation of oxide scales on the surface by exhausting of aluminum and chromium from the alloy inside and formation of an area of dissolved γ´ precipitates.

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

**Figure 23.** *(a) Oxide scales after 8 cycles (b) Oxide scales after 18 cycles.*

**Figure 24.** *Thicknesses of oxides and gama-solid solution after 8, 18, 28 and 38 cycles of testing.*

Next cyclic oxidation cycles resulted in linearly increasing thicknesses of oxide scales and area of dissolved precipitates, i.e., solid solution γ without strengthening γ´-phase. The γ-solid solution grown about 5 times faster than oxide scales (**Figure 24**). The continuously growth of oxides and their mass gain on the surface of samples without coating has shown that although the oxides underwent spallation processes during each cooling cycle of testing, the surface of alloy was doped by aluminum from the bulk after each cycle (see **Figure 14**), whereas strength and creep properties of the alloy decreased. Decreasing mass gain started from the 29th cycle when the forming of oxides continued mainly in the inside of the alloy whereas the surface was subjected to the spalling process. X-ray diffraction analysis of phase of the sample after 38 cycles of cyclic oxidation revealed a presence of protective oxide phases on the base of Al and Ti in the mixture with NiAl2O4 spinel phase (**Figure 25**). NiAl2O4 could be formed in the process of cyclic oxidation by the reaction of NiO and Al2O3.

**Figure 25.** *X-ray analysis of phases on the surface of sample after 38 cycles of cyclic oxidation.*

## *2.2.3.2 Microstructural examination of IN 713 LC with Al coating*

**Figure 26a** and **b** represent the microstructure of superalloy IN 713 LC with Al coating deposited by the "out-of-pack" method before the cyclic oxidation test. The upper layer of the coating was composed from the NiAl solid solution and the inner one contained Ni3Al solid solution with particles of Cr, Nb, Ti and Mo. The thicknesses of upper and inner sublayers were 32.8 μm and 17.7 μm, respectively. **Figures 27**–**30** show changes of microstructures of samples after 8, 18 and 28 cycle, respectively. They revealed that the degradation of the protective coating continued very fast and a zero gain was reached after 28 cycles of testing (see **Figure 14**). This mass change of 0 mg/cm<sup>2</sup> compared to the original state of sample indicated that the reserve of aluminum was exhausted and the protected function of the coating was stopped. The oxide thicknesses on the sample surfaces after 8, 18 and 28 cycles were 6.68, 4.97 and 4.68 μm, respectively.

The oxidation behavior at the beginning of the test was controlled by the formation of a fast-growing layer of a mixed oxide. At cyclic oxidation times longer than

*(a) Microstructure of IN 713 LC with Al coating before testing (b) Detail on the coating before testing.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 27.**

*Microstructure of IN 713 LC with Al coating after 8 cycles of testing.*

## **Figure 28.**

*Microstructure of IN 713 LC with Al coating after 18 cycles of testing.*

12 cycles this rather thick layer started to spall off. In accordance with the unit mass change (**Figure 14**) up to 12th cycle the protective oxide scales formed on the surface of samples prevailed over the effects of spallation and starting from the 13th cycle the

#### **Figure 30.**

*Intermediate zone between the inner layer of coating and the superalloy of sample after 28 cycles of testing.*

#### **Figure 31.** *Crack in the Al coating after 8 cycles of testing.*

spalling processes become dominating. Both the weight gain in **Figure 14** and the measured oxide layer thickness demonstrate this behavior. A crack in the coating of sample after 8 cycles was found as a result of thermal stresses (**Figure 31**). A major problem of such coatings is that the coefficient of thermal expansion of the alumina layer differs from the coefficient of expansion of the base material.

During thermal cycling, stresses arise between the aluminum oxide top layer and the coating material. The resulting oxide layer is relatively brittle and tends to crack and peel off, exposing the fresh surface to a damaging atmosphere. This repeating process consumes the aluminum in the coating. When the aluminum level in the coating drops below a certain point, the coating becomes ineffective as an alumina generator and the protective benefits of the coating material are lost.

Process of spalling can be examined by the view on the surfaces of samples after 8, 18 and 28 cycles (**Figure 32a**–**d**).

Dark areas in **Figure 32b–d** responded to protective alumina scales and white places to the Ni3Al phase with an amount of Cr, Nb, Ti and Mo. This is in accordance with EDS analyses of elements from surfaces and cross sections of each of samples. An example of this examination from the sample after 18 cycles we can see in **Figure 33** where phase 1 corresponds to the oxides and phase 2 to the solid solution Ni3Al.

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 32.**

*(a) The surface of IN 713 LC with Al coating before testing (b) The surface of IN 713 LC with Al coating after 8 cycles of testing (c) The surface of IN 713 LC with Al coating after 18 cycles of testing and (d) The surface of IN 713 LC with Al coating after 28 cycles of testing.*


**Figure 33.**

*EDS analyses of phases from surface and cross section of samples after 18 cycles of testing.*

## *2.2.3.3 Microstructural examination of IN 713 LC with AlSi coating*

Si modified aluminide coatings were made on the surface of IN 713 LC samples by the method of "pack-cementation". Silicon was added to the aluminide coatings to improve their oxidation resistance and the oxide scale adherence and, as a consequence, the oxidation rate was lower. After the pack cementation process some of Ni atoms were replaced by Si in the solid solution of aluminide phase. This resulted from the X–ray diffraction phase analysis (**Figure 34**) since no Sicontaining phases were found. During pack aluminizing, the superalloy samples to be coated were placed in an air-tight retort containing a mixture powder of aluminum and silicon activated with ammonium chloride and an inert Al2O3 filler which prevented the sources form sintering. The box was then inserted into a furnace and heated in a protective atmosphere. The pack cementation process is essentially an in situ chemical vapor deposition (CVD) coating process. The coating achieved by this method and subsequent heat treatment was composed from outer and small inner sublayers (**Figures 35** and **36**). According to the results of the EDS analyses, the element distribution of outer and inner coating layers and the substrate is

#### **Figure 34.**

*The X-ray diffraction phase analysis from the surface of the sample IN 713 LC with the AlSi coating.*

#### **Figure 35.**

*The cross section of sample IN 713 LC with AlSi coating before cycling oxidation test.*


*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Table 2.**

*Element distribution in the IN 713 LC superalloy with the AlSi coating in at. %.*

displayed in **Table 2**. The element distribution confirmed that the outer layer contained a large amount of small precipitates containing atoms with higher atomic numbers such as Cr, Ti and Mo which are present in the substrate. This is a typical feature for inward diffusion of Al and Si. The small diffusion underneath suggests a limited degree of the outward diffusion of Ni. The total thickness of the coating including the diffusion zone was approximately 77 μm. The X-ray diffraction phase analysis by JEOL JDX-7S from the sample surface revealed that the matrix of the coating was the Ni2Al3 phase in which particles of niobium aluminide were present (**Figure 34**).

Microstructures of IN 713 LC samples with AlSi coating after testing of cyclic oxidation are visible in **Figures 37**–**41**. Specimens after 8 cycles of testing changed

**Figure 37.** *(a) Microstructure after 8 cycles (b) Crack in the coating after 8 cycles.*

**Figure 38.** *Microstructure after 18 cycles.*

**Figure 39.** *(a) Microstructure after 28 cycles (b) Cracks after 28 cycles.*

their microstructural view. The white particles visible in **Figure 37a** are composed from elements as Si, Cr, Ti, Nb and Mo as found on the basis of the EDX analyses. The content of Al in the coating was about 50 at. % and after 8 cycles decreased to 29 at. % *Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

**Figure 40.** *Microstructure after 38 cycles.*

**Figure 41.** *View on the thicknesses of coatings after 8, 18, 28 and 38 cycles of testing.*

and then remained constant up to 38 cycles. The content of Ni exhibited an opposite behavior and increased from the original 38 at. % to 64 at. % after 8 cycles and then remained constant up to 38 cycles.

The content of particles based on Si, Ti, Cr, Nb and Mo maintained practically identical during all the testing cycles. As one can see from **Figure 41**, the coating thickness grew larger after 8 cycles of testing and then remained practically unaffected up to 38 cycles.

There were practically no changes in the thickness of surface oxide scales during the cycling. **Figure 42a–h** represent surfaces of coatings before and after 8, 18, 28 and 38 cycles of testing. Contents of Al and Ni before testing changed from about 51 at. % and 40 at. % to 92,5 at. % and 5,8 at. % after testing, respectively. Alumina phase on the surface was subjected to the processes of buckling and scaling off and white particles were exposed. A large amount of small cracks on the top of coating (**Figure 42c, e, g**) represents an evidence of such a mechanism. Since the concentration of white phases during the cycling did not change too much it means that new alumina phases were created simultaneously with the scales. This behavior along with the small mass gain to the 11th cycle of testing and a very small weight loss up to the end of the cycling (**Figure 14**) confirm a good resistance of IN 713 LC with AlSi coating to the thermal cycling.

**Figure 42.** *Surfaces of Inconel 713LC samples before cyclic oxidation and after 8, 18, 28 and 38 cycles of testing.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*


#### *2.2.3.4 Microstructural examination of MAR 247 LC without coating*

MAR 247 LC belongs to widely used high temperature nickel-base superalloy. It is a cast polycrystalline material used especially for turbine blades and discs. The chemical composition (**Table 1**) of the alloy has, in comparison with IN 713 LC, a larger content of heavy elements as tungsten and tantalum which act as the most efficient matrix hardeners and promote creep strength. Cobalt in MAR 247 LC has only a small direct influence on strengthening but it affects the solubility of elements in the matrix solid solution and often raises the temperature of solidus which can lead to the larger amount of precipitates at low to medium temperatures [28]. Carbon is present at higher concentration (0.15 wt.%) than in the IN 713 LC (0.04 wt.%) and it combines with reactive elements such as titanium, tantalum, hafnium and tungsten to form MC carbides. During processing or service these carbides can decompose to other forms as M23C6 and M6C which are rich in chromium, molybdenum and tungsten. Excessive amount of tungsten together with molybdenum and chromium leads to the formation of so-called topologically close-packed (TCP) phases. Various semi-empirical models are used for balancing composition of superalloys to avoid forming of these undesirable phases. The material before testing and without coating was subjected to a hot isostatic pressing and, subsequently, it was heat treated by two steps to reach the microstructure shown in **Figure 43a**. This initial microstructure of the MAR 247 LC alloy consisted of γ Ni-rich solid solution containing a dispersion of γ<sup>0</sup> precipitates, carbide particles, and γ/γ<sup>0</sup> eutectics as depicted in **Figure 43b**.

As one could see in **Figure 15**, the superalloy MAR 247 LC without coating revealed a very poor resistance to thermal cycling. Mass loss started immediately after the 1st cycle and continued during subsequent 12 cycles when test was stopped. Microstructures of this alloy after 8 cycles of testing well reflect that behavior – see **Figures 44** and **45**.

Precipitation-hardenable superalloys usually have a good oxidation resistance in oxidizing atmospheres within their normal range of service temperatures. Exposure to high-temperature environments can cause changes in the alloy composition near the surface. **Figure 45** shows the changes in the subsurface microstructure formed under a high temperature environment. As we can see in **Figure 45**, the surface of the alloy is covered by oxidation scales based on

#### **Figure 43.**

*(a)View on the microstructure of MAR 247 LC without coating before testing (b) Phases in the microstructure of the MAR 247 LC alloy.*

#### **Figure 44.** *View on the structure of MAR 247 LC after 8 cycles.*

aluminum (area 5 in **Figure 45**) but an internal oxidation is also visible. Preferential oxygen attacks on carbide phases were observed. Since certain elements as aluminum or chromium are consumed by the scale layer, the bulk composition can become depleted. Subsurface zone (area 2 in **Figure 45**) is composed from a solid solution of elements in nickel with small amounts of aluminum and without γ´ coherent precipitates. Chemical content of phases highlighted in the **Figure 45** is in **Table 3**.

A comparison of both surfaces, i.e. before and after cycling oxidation, is presented in **Figures 46**–**48**. The EDX analyses confirmed that the chemical composition of the alloy on the sample surface (**Figure 46**) without cycling is identical with that declared in **Table 1**. The specimen surface after 8 cycles is covered by oxidation scales as can be seen in **Figures 47** and **48**. The surface is mainly rich in aluminum, chromium and hafnium. Tungsten and tantalum are present in carbides

## **Figure 45.**

*Subsurface zone of the sample after 8 cycles.*


#### **Table 3.**

*Element distribution in MAR 247 LC without coating after 8 cycles of testing in at. %.*

(white particles in **Figures 47 and 48**). The detail depicted in **Figure 48** shows an attack of carbides along grain boundaries even more clearly than one could see from **Figure 45**.

**Figure 47.**

*Surface of MAR 247 LC without coating without coating before testing after 8 cycles of testing.*

**Figure 48.** *Oxide scales on the surface of MAR 247 LC without coating after 8 cycles of testing.*

## *2.2.3.5 Microstructural examination of MAR 247 LC with Al coating*

**Figures 49** and **50** show cross-sectional images of the sample MAR 247 LC with Al coating before testing of cycle oxidation at 1100°C in the air environment. The twolayered coating (**Figure 50**) includes the upper NiAl layer which serves as aluminum reservoir and the inner diffusion layer of solid solution of Ni with a smaller content of Al and carbide particles on the basis of Cr, W, Hf and Mo. The chemical content of elements in the highlighted areas (1-4) in **Figure 50** achieved from the EDX analysis is in **Table 4**. The average thicknesses of upper and inner parts are 24.9 μm and 18.2 μm, respectively.

The view on the microstructure of the specimen after 8 cycles of testing at 1100°C in the air for 8x23 hours is in **Figure 51**. The surface is covered by an oxide film and, with respect to this diffusion process, the upper part of the coating is depleted of aluminum. The content of aluminum in the matrix of upper layer dropped from 49.5 at. % before testing to 27.4 at. % after 8 cycles of oxidation. In this zone one can observe grains with a higher concentration of nickel (65 at. %) than in the matrix (in **Figures 51**–**53** described as "Grains of Ni"). These grains are bigger in samples after 18

#### **Figure 49.**

*SEM micrograph of the microstructure of the sample MAR 247 LC with Al coating before testing at low magnification.*

#### **Figure 50.**

*Cross-section of sample MAR 247 LC with Al coating before cycle oxidation testing at 1100°C.*


#### **Table 4.**

*Element distribution in MAR 247 LC with Al coating before testing in at. %.*

and 28 cycles of oxidation. The microstructure view on the sample after 38 cycles (**Figure 54**) is different from the previous ones.

Coating upper layer of the sample after 38 cycles has two phases as in previous samples (**Figures 51**–**53**) but proportion of these phases is changed. Grains of Ni are extended to a large extent that they can be assumed as matrix of the sample after 38 cycles and second dark phases mean unchanged grains of the upper layer matrix of samples after 8, 18 and 28 cycles.

Moreover, needles of topologically close packed phases (TCP) were observed. Presence of refractory elements in the superalloys provide strength benefits from solid solution hardening but a tendency for an alloy instability due to a formation of TCP phases (or secondary reaction zone phenomena) is high. The formation of these phases has a detrimental effect because of their brittle nature and depletion of Ni-rich matrix from strengthening elements [29]. The thickness of coating after 8 cycles rose up twice in comparison with original state and then it kept changeless up to the end of thermal cycling.

**Figure 51.** *Microstructure of MAR 247 LC.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

**Figure 53.** *Microstructure of MAR 247 LC.*

#### **Figure 54.**

*Microstructure of MAR 247 LC with Al coating after 28 cycles with Al coating after 38 cycles.*

Appearance of the sample surface after 8 cycles of oxidation is in **Figure 55** where a spallation of small oxide segments can be seen. Similar images were observed on samples after 18, 28 and 38 cycles (**Figure 56**).

The spallation started already at very early stages of oxidation. Stress generation within the oxide during its growth and its release due to cracking in the scale or creep of the substrate metal led to the spallation process and exposure of the substrate. In spite of the loss of protective scales during the exposure by a substantial number of 38 cycles some new scales were reformed. On the basis of data achieved from the thermogravimetric analyses (**Figure 15**) the MAR 247 LC with Al coating after 38 cycles of cyclic oxidation can be still considered to be a good protective system.

## *2.2.3.6 Microstructural examination of MAR 247 LC with AlSi coating*

Pack powder mixture for codepositing Al and Si on MAR 247 LC superalloys by the pack cementation process was used to form AlSi diffusion coating. Silicon (similarly to

**Figure 55.** *General view on the surface of sample after 8 cycles of oxidation and detail from spalled area.*

**Figure 56.** *Spalled area and cracks on the oxide surface of sample after 38 cycles.*

chromium) was added to the aluminide coating to enhance the resistance against oxidation and sulfur-carrying gases i.e. hot corrosion. However, application of Si is limited to small amounts because silicon bears the risk of forming low-melting phases in nickel-base superalloys [23]. Overall microstructure view on the sample MAR 247 LC with AlSi coating is visible in **Figure 57** where coating is two-layered. The upper

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 57.**

*General cross-section view on the sample MAR 247 LC with AlSi coating before testing.*

#### **Figure 58.**

*The coating at higher magnification.*


#### **Table 5.**

*Element distribution in MAR 247 LC with AlSi coating before testing in at. %.*

part has average thickness of 94.4 μm and the inner diffusion part is much smaller. Detailed chemical compositions achieved by EDX analyses of individual marked zones (see **Figure 58**) is in **Table 5**.

Silicon is the most placed in region 1 and it has thickness about 12.5 μm. Lower regions 2 and 3 have hardly any silicon and chemical content of others elements is similar to the region 1. The upper part of this coating has a few of precipitates based on Si and refractory elements (point 6 in **Figure 58**). Thin diffusion coating part (area 4 in **Figure 58**) is formed from NiAl solid solution with precipitates on the base of refractory elements and its chemical content is very similar to the diffusion zone of Mar 247 LC with Al coating (see **Table 4**). EDX analyses show that representation of silicon gradually falls from the surface to the diffusion zone of the coating. Under this coating, we can observe dendritic microstructure of the superalloy. Ni-rich γ matrix with γ´ semi-coherent precipitates contains MC carbides characterized by different shapes (from discrete blocky precipitates of a diversified shape and size to the shape known as Chinese script) and borides (**Figure 57**). MC carbides precipitated in the final stage of solidification via eutectic reaction with *γ*-matrix in the interdendritic areas. The secondary M23C6 carbides are very fine and exist at the grain boundaries.

Testing of samples on cyclic oxidation at the temperature 1100°C in the air environment showed that this system MAR 247 LC superalloy/AlSi coating had excellent resistance to cyclic oxidation. The amount of aluminum in the upper part of coating decreased after first cycles on the average from 47 at. % to 26 at. % and then up to 38 cycles of testing remained at the same value. This means that aluminum reservoir is sufficient to form protective oxides on the surface and the oxidation rate is slow. From the point of view of engineering design, kinetics of oxidation is very important because it gives an estimate of design life of system superalloy/coating. The microstructures of samples after 8, 18, 28 and 38 cycles are visible in **Figures 59**–**62**. The thickness of the coating of sample after 8 cycles rose sharply 3.3 times and then the thicknesses of all tested samples remained the same. Topologically close packed phases emerged after 8 cycles of testing, their amounts gradually increased to 38 cycles. This behavior was although observed in samples of MAR M247 LC with Al coating. From the point of view of other phases, the microstructure remained without significant changes up to 38 cycles of testing of the cycle oxidation. Chemical contents of phases found in MAR 247 LC with AlSi coating after 38 cycles were similar to the phases in MAR 247 LC with Al coating.

**Figure 59.** *Microstructure of MAR 247 LC.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 60.**

*Microstructure of MAR 247 LC with AlSi coating after 8 cycles with AlSi coating after 18 cycles.*

**Figure 61.** *Microstructure of MAR 247 LC.*

#### **Figure 62.**

*Microstructure of MAR 247 LC with AlSi coating after 28 cycles with AlSi coating after 38 cycles.*

Surfaces of MAR 247 LC with AlSi coating samples after 8, 18, 28 and 38 cycles of testing represent **Figures 63**–**69**. We can observe that protective alumina oxide after 8 cycles spalled out and the layer composed from solid solution of Al in Ni and precipitates based on refractory elements revealed (**Figure 64**). Some of these spalled

**Figure 63.** *Surface of MAR 247 LC with AlSi coating after 8 cycles (left BSE image, right SE image).*

**Figure 64.** *Detail from the surface displayed in Figure 63 (left BSE image, right SE image).*

**Figure 65.** *Surface of MAR 247 LC with AlSi coating after 18 cycles (left BSE image, right SE image).*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

#### **Figure 66.**

*Surface of MAR 247 LC with AlSi coating after 28 cycles.*

### **Figure 68.**

*Surface of MAR 247 LC with AlSi coating after 38 cycles.*

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

parts had a bigger size (the white particle in **Figure 64**). Next cyclic oxidation caused new alumina scales formation and white particles seen in **Figure 65** were smaller. Creation of protective alumina scales continued (see **Figure 66**) and after 38 cycles of testing protective oxides (**Figure 68**) covered almost all the surface of coated superalloy.

## **Author details**

Marta Kianicová Alexander Dubcek University in Trencin, Trencin, Slovakia

\*Address all correspondence to: marta.kianicova@tnuni.sk

© 2022 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

## **References**

[1] Singh R. Applied Welding Engineering. In: Processes, Codes and Standards. 3rd ed. BH Amsterdam: Elsevier Inc.; 2020. p. 442. DOI: 10.1016/ C2019-0-03490-5

[2] Xiongfeng H, Siyu J, Fuqiang L, Li J, Xiaoqiang L, Shengguan Q. Investigation on the parameters optimization and sliding wear behaviors under starved lubrication of discrete laser surface hardened 25CrNi2MoV steel. Tribology International. 2021;**163**:163. DOI: 10.1016/j.triboint.2021.107176

[3] Zhang B, Chen J, Wang P, Sun B, Cao Y. Enhanced strength-ductility of CoCrFeMnNi high-entropy alloy with inverse gradient-grained structure prepared by laser surface heat-treatment technique. Journal of Materials Science & Technology. 2022;**111**:111-119. DOI: 10.1016/j.jmst.2021.09.043

[4] Laketić B, Rakin M, Momčilović M, Ciganović J, Veljović D, Cvijović-Alagić I. Influence of laser irradiation parameters on the ultrafine-grained Ti–45Nb alloy surface characteristics. Surface and Coatings Technology. 2021;**418**:14. DOI: 10.1016/j.surfcoat.2021.127255

[5] Guo Y, Jia J, Kong B, Peng H, Zhang H. Heat treatment induced phase transition and microstructural evolution in electron beam surface melted Nb-Si based alloys. Applied Surface Science. 2017;**423**:417-420. DOI: 10.1016/ j.apsusc.2017.05.248

[6] Vazquez-Martinez J, Del Sol II, Arrien EU, Batista M, Salguero J. Laser surface texturing as a finishing process for aerospace alloys. In: Handbooks in Advanced Machining and Finishing. 1st ed. BH Amsterdam: Elsevier Inc.; 2021. DOI: 10.1016/B978-0-12-817452- 4.00010-5

[7] Muthukumaran G, Dinesh BP. Metallurgical characterization of laser hardened, mechanically textured 2.5 Ni-Cr-Mo low alloy steel and optimization using RSM. Optics & Laser Technology. 2021;**141**. DOI: 10.1016/j.optlastec. 2021.107126

[8] Zaleski B. A study on the properties of surface – active fluids sused in burnishing and shot peening processes. Advances in Science and Technology Research Journal. 2016;**10**:235-239. DOI: 10.12913/22998624/64012

[9] Maleki E, Unal O, Guagliano M, Bagherifard S. The effects of shot peening, laser shock peening and ultrasonic nanocrystal surface modification on the fatigue strength of Inconel 718. Materials Science and Engineering: A. 2021;**810**:11. DOI: 10.1016/j.msea.2021.141029

[10] Maleki E, Farrahi GH, Kashyzadeh KR, Unal O, Gugaliano M, Bagherifard S. Effects of conventional and severe shot peening on residual stress and fatigue strength of steel AISI 1060 and residual stress relaxation due to fatigue loading: experimental and numerical simulation. Metals and Materials International. 2021;**27**: 2575-2591. DOI: 10.1007/s12540-020- 00890-8

[11] Agaram S, Srinivasan SM, Kanjarla AK. Crystal plasticity modelling of stability of residual stresses induced by shot peening. International Journal of Mechanical Sciences. 2022;**230**:13. DOI: 10.1016/j.ijmecsci.2022.107526

[12] Fu P, Su H, Li Z, Dai P, Tang Q. Enhancing mechanical properties of dual-phase Al0.5CoCrFeNiSi0.25 high entropy alloy via thermomechanical treatment. Journal of Alloys and

*Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

Compounds. 2022;**921**:11. DOI: 10.1016/ j.jallcom

[13] Wang X, Xiao Z, Qiu W, Li Z, Liu F. The evolution of microstructure and properties of a Cu–Ti–Cr–Mg–Si alloy with high strength during the multistage thermomechanical treatment. Materials Science and Engineering: A. 2021;**803**:9. DOI: 10.1016/j.msea.2020. 140510

[14] Mittemeijer EJ, Somers MAJ. Kinetics of thermochemical surface treatments. Thermochemical Surface Engineering of Steels. Elsevier Inc.; 2015. 113-140 p. DOI: 10.1533/9780857096524. 1.113

[15] Yao J et al. On the growth of functionally graded self-lubricating layer during a plasma-assisted thermochemical treatment of M50NiL steel. Applied Surface Science. 2022;**584**: 10. DOI: 10.1016/j.apsusc.2022.152517

[16] Prasad MA, Dharmalingam G, Salunkhe S. Microstructural evaluation of gas nitrided AISI 316 LN austenitic stainless steel. Journal of the Energy Institute. 2022;**104**:112-123. DOI: 10.1016/j.joei.2022.07.015

[17] Naeem N et al. Improved wear resistance of AISI-1045 steel by hybrid treatment of plasma nitriding and postoxidation. Tribology International. 2022; **175**:11. DOI: 10.1016/j.triboint.2022. 107869

[18] Chen W et al. The thermal process for hardening the nitrocarburized layers of a low-carbon steel. Scripta Materialia. 2022;**210**:4. DOI: 10.1016/j.scriptamat. 2021.114467

[19] Katsamas AI, Haidemenopoulos GN. Laser-beam carburizing of low-alloy steels. Surface and Coatings Technology. 2001;**139**:183-191. DOI: 10.1016/ S0257-8972(00)01061-6

[20] Kucherov FA, Romashov LV, Anamikov VP. Development of 3D+G printing for the design of customizable flow reactors. Chemical Engineering Journal**.** 2022;**430**:9. DOI: 10.1016/j.cej. 2021.132670

[21] Kalkal A et al. Recent advances in 3D printing technologies for wearable (bio) sensors. Additive Manufacturing. 2021; **46**:20. DOI: 10.1016/j.addma.2021. 102088

[22] Wang W, Cui W, Xiao Z, Qin G. The improved corrosion and wear properties of Ti-Zr based alloys with oxide coating in simulated seawater environment. Surface and Coatings Technology. 2022; **439**:11. DOI: 10.1016/j.surfcoat.2022. 128415

[23] Welegergs GC, Akoba R, Sacky J, Nuru ZY. Structural and optical properties of copper oxide (CuO) nanocoatings as selective solar absorber. Materials Today: Proceedings. 2021;**36**: 509-513. DOI: 10.1016/j.matpr.2020. 05.298

[24] Dai D, Zhou D, He L, Wang C, Zhang C. Graphene oxide nanocoating for enhanced corrosion resistance, wear resistance and antibacterial activity of nickel-titanium shape memory alloy. Surface and Coatings Technology. 2022; **431**:13. DOI: 10.1016/j.surfcoat.2021. 128012

[25] Li M, Han B, Song L, He Q. Enhanced surface layers by laser cladding and ion sulfurization processing towards improved wear-resistance and self-lubrication performances. Applied Surface Science. 2020;**503**:11. DOI: 10.1016/j.apsusc.2019.144226

[26] He X, Wu J, Chen Y, Zhang L, Sheng X. A trace amount of

MXene@PDA nanosheets for lowtemperature zinc phosphating coatings with superb corrosion resistance. Applied Surface Science. 2022;**603**:11. DOI: 10.1016/j.apsusc.2022.154455

[27] Tian Y et al. Accelerated formation of zinc phosphate coatings with enhanced corrosion resistance on carbon steel by introducing a-zirconium phosphate. Journal of Alloys and Compounds. 2020;**831**:9. DOI: 10.1016/ j.jallcom.2020.154906

[28] Vazirgiantzikis I, George SL, Pichon L. Surface characterisation and silver release from Ti-6Al-4V and anodic TiO2 after surface modification by ion implantation. Surface and Coatings Technology. 2022;**433**:10. DOI: 10.1016/ j.surfcoat.2021.128115

[29] Ranjan R, Das AK. Protection from corrosion and wear by different weld cladding techniques: A review. Materials Today: Proceedings. 2022;**57** (3):1687-1693 DOI: 10.1016/j.matpr. 2021.12.329

[30] Qian Y et al. Microstructure and mechanical properties of SiC particle reinforced Zr-based metallic glass surface composite layers produced by laser alloying. Surface and Coatings Technology. 2022;**446**:10. DOI: 10.1016/ j.surfcoat.2021.128784

[31] Monisha K et al. Titanium boride and titanium silicide phase formation by high power diode laser alloying of B4C and SiC particles with Ti: Microstructure, hardness and wear studies. Materials Today. Communications. 2022;**31**:19. DOI: 10.1016/j.mtcomm.2022.103741

[32] Yang GJ, Suo XK, Li GR. Advanced Nanomaterials and Coatings by Thermal Spray. 1st ed. BH Amsterdam: Elsevier Inc.; 2019:328. DOI: 10.1016/C2017-0- 00237-9

[33] Hu H et al. Wear-resistant ceramic coatings deposited by liquid thermal spraying. Ceramics International. 2022; **48**(22):33245-33255. DOI: 10.1016/j. ceramint.2022.07.267

[34] Poursaee A, editor. Corrosion of Steel in Concrete Structures. 1st ed. Cambridge UK: Elsevier Inc.; 2016. p. 285. DOI: 10.1016/C2014-0-01384-6

[35] Wei X et al. Deposition of DLC films on the inner wall of U-type pipes by hollow cathode PECVD. Diamond & Related Materials. 2021;**114**:11. DOI: 10.1016/j.diamond.2021.108308

[36] Biava G et al. Evaluation of high temperature corrosion resistance of CrN, AlCrN, and TiAlN arc evaporation PVD coatings deposited on Waspaloy. Surface and Coatings Technology. 2022;**348**:15. DOI: 10.1016/j.surfcoat.2022.128398

[37] Agüero A, Østergård MJL, Hansson AN, Gutierrez M. Thermal cyclic resistance and long term interdiffusion properties of slurry aluminide coatings modified with Si. Results in Surfaces and Interfaces. 2022;**6**:16. DOI: 10.1016/j.rsurfi.2022.100042

[38] Pokluda J, Kianicová M. Gas turbines. In: Injeti G, editor. Damage and Performance Assessment of Protective Coatings on Turbine Blades. Rijeka Croatia: IntechOpen; 2010. DOI: 10.5772/ 45608

[39] Kianicová M, Kafrík J, Trník J. Degradation of aluminide coatings deposited on nickel superalloys. Procedia Engineering. 2016;**136**:346-352. DOI: 10.1016/j.proeng.2016.01.221

[40] Fu C, Wang C, Ren ZM, Cao GH. Comparison of microstructure and oxidation behavior between Pt-free and Pt-modified -Ni Si coatings on Ni-based *Cyclic Oxidation of Diffusion Aluminide Coatings DOI: http://dx.doi.org/10.5772/intechopen.107972*

superalloys. Corrosion Science. 2015;**98**: 211-222. DOI: 10.3323/jcorr1991.50.582

[41] Takahasi RJ, Assis JMK, Neto FP, Reis DAP. Thermal conductivity study of ZrO2-YO1.5-NbO2.5.TBC. Journal of Materials Research and Technology. 2022;**19**:4932-4938. DOI: 10.1016/j.jmrt. 2022.07.037

[42] Liew W et al. Thermal stability, mechanical properties, and tribological performance of TiAlXN coatings: Understanding the effects of alloying additions. Journal of Materials Research and Technology. 2022;**17**:961-1012. DOI: 10.1016/j.jmrt.2022.01.005

[43] Pakseresht A, Sharifahmadian O. Handbook of Research on Tribology in Coatings and Surface Treatment. 1st ed. Hersey, Pennsylvania, USA: IGI Global; 2022. p. 470. DOI: 10.4018/978-1- 7998-9683-8

[44] Zare Mohazabie MS, Nogorani FS. The addition of zirconium to aluminide coatings: The effect of the aluminide growth mode. Surface and Coatings Technology. 2019;**378**:8. DOI: 10.1016/ j.surfcoat.2019.125066

[45] Rahmani KH, Nategh S. Isothermal LCF behavior in aluminide diffusion coated René 80 near the DBTT. Materials & Design. 2009;**30**:1183-1192. DOI: 10.1016/j.matdes.2008.06.064

[46] Peng X. 3-Metallic coatings for hightemperature oxidation resistance. In: Xu H, Guo H, editors. Thermal Barrier Coatings. A volume in Woodhead Publishing Series in Metals and Surface Engineering. 2011. p. 339. DOI: 10.1533/ 9780857090829.1.53

[47] Bozza F et al. Diffusion mechanisms and microstructure development in pack aluminizing of Ni-based alloys. Surface and Coatings Technology. 2014;

**239**:147-159. DOI: 10.1016/j.surfcoat. 2013.11.034

[48] Patent US8137749B2, SNECMA, 2012, Method of aluminization in the vapour phase on hollow metal parts of a turbomachine

[49] Squillace A, Bonetti R, Archer NJ, Yeatman JA. The control of composition and structure of aluminide layers formed by vapour aluminizing. Surface and Coatings Technology. 1999;**120-121**: 118-123. DOI: 10.1016/S0257-8972(99) 00347-3

[50] Grégoire B, Bonnet G, Pedraza F. Development of a new slurry coating design for the surface protection of gas turbine components. Surface and Coatings Technology. 2019;**374**:521-530. DOI: 10.1016/j.surfcoat.2019.06.020

[51] Grégoire B, Oskay C, Meisner T, Galetz M. Corrosion performance of slurry aluminide coatings in molten NaCl–KCl. Solar Energy Materials and Solar Cells. 2021;**223**:20. DOI: 10.1016/j. solmat.2021.110974

[52] Xu C, Gao W. Pilling-Bedworth ratio for oxidation of alloys. Materials Research Innovations. 2000;**3**:231-235. DOI: 10.1007/s100190050008

[53] Li Y, Li Z, Liu L, Cai N. Measuring the fast oxidation kinetics of a manganese oxygen carrier using microfluidized bed thermogravimetric analysis. Chemical Engineering Journal. 2020;**385**:12. DOI: 10.1016/j.cej.2019. 123970

[54] Young DJ. In: Young DJ, editor. High Temperature Oxidation and Corrosion of Metals. BH Amsterdam: Elsevier Inc.; 2016. p. 733. DOI: 10.1016/C2014-0- 00259-6

## **Chapter 6**
