Surface Hardening Of Steel

Edit this page (last edited April 9, 2003)
Frequently Asked Questions | Title Index | Recent Changes | User Preferences | Sand Box | Help

Surface Hardening of Steel

There are many different types of heat-treatment process used to modify the surface properties of steel components. The majority of these processes are used to produce harder, more wear and fatigue resistant surfaces than could be obtained from the base material. A component that has a tough and relatively ductile core with a hard wear resistant case is the result of heat treatment and this approach has many attractions from an engineering point of view.

Surface treatments fall into two main categories:
  1. Thermochemical Processes such as carburising, cabonitriding, nitriding and ferritic nitro-carburising. In these processes the local chemistry of the steel is modified by the introduction of carbon and/or nitrogen and occasionally other elements such as boron.
  2. Localised Thermal Treatment such as induction hardening, flame hardening, laser hardening and electron beam hardening. In these processes selective surface heating is used to allow local hardening to take place whilst leaving the core properties of the component unaffected.

There are many variations of the basic processes not only due to the demands of engineers needing to meet specific service requirements but also due to commercial competition.

The different approaches to thermochemical processing are usually classified according to the method of heat treatment and can be classified as follows:

  1. Gaseous Processes e.g. gas carburising, gas nitriding
  2. Salt-Bath Processes e.g. carbonitriding (also known as cyanide hardening)
  3. Vacuum Processes e.g. vacuum carburising, ion nitriding, plasma nitriding)
  4. Pack Processes e.g. pack carburising, metallising

The features of these processes that are of most importance are:
  1. Processing Temperature, since this affects distortion
  2. Mechanical Properties of the case particularly hardness
  3. The depth of the case depending on the application
  4. The service behaviour of the case-core composite.

One interesting point to note is that distortion is mostly affected by process temperature. The composition of the steel used has little effect on distortion but it does affect the choice of quenching media, which can have a significant effect on distortion.

Process Temperatures.

Distortion is directly related to process temperature and this parameter also influences case depth. There are two basic types of process:
  1. Those carried out at low temperature (<600 Celsius) when most steels are in a ferritic condition.
  2. Those carried out at higher temperatures (>800 Celsius) when most steels are in the austenitic condition.

The ferritic condition is the stable form of carbon and low alloy steels below about 720 Celsius. If carbon steels are heated above this temperature austenite begins to form. As the temperature rises increasing amounts of austenite are formed until about 800 Celsius (for 0.4% Carbon Steels) when the entire structure is austenitic. The solubility of carbon is much higher in austenite than in ferrite and carbon atoms diffuse through the crystal matrix and take up different positions within the crystals. Ferrite structures are Body Centred Cubic (BCC) crystals and austenite is a Face Centred Cubic Crystal (FCC) and the change in structure starts at about 720 Celsius. There are some changes in these temperatures depending on alloy content.(For example 18/8 Stainless Steel is Austenitic right down to room temperature) The reason that this reaction is important as it is this fundamental allotropic behaviour that allows steel to be hardened and tempered. If steel is heated into the austenite region a single phase FCC structure will result. If it is then cooled slowly the resulting structure at room temperature will be a relatively soft mixture of ferrite and a new carbon-rich component known as pearlite. If the same material is rapidly cooled into oil or water the austenite, which has carbon atoms sitting in specific sites within the crystal lattice will not have time to transform properly to the low temperature BCC structure as the transformation needs time for diffusion. The result is a heavily distorted structure which is known as martensite. Martensite is extremely hard and very brittle (the only fully martensitic component that I know of is a file !) and components are normally tempered to reduce the level of residual stress and make the component more ductile and tougher. The higher the percentage carbon in a steel (Up to about 0.8%) the higher the hardness that can be produced by quenching. Alloy additions are used to modify diffusion rates of carbon atoms and the depth of hardness that can be produced will depend on this alloy content. ( Significant depth of hardness could be obtained by extremely rapid cooling but this would cause significant distortion and probably cracking due to thermal shock). The addition of carbon also increases the strength of a steel but unfortunately as strength increases toughness and fatigue resistance tend to reduce. (This is an over-simplification but broadly true). Fatigue resistance is vitally important as 95% of all component failures are caused by metal fatigue of one sort or another.

So why is this basic knowledge important? If many of the components used in motor cars were not surface treated they would either wear out extremely quickly or they would be so brittle that they could shatter under extreme conditions. The development of reliable and durable motor cars has depended very significantly on the ability to successfully modify local properties.

Typical Processes


This is one of the most basic processes used in the surface treatment of steels. Virtually every gear produced for a motor car is carburised using one of the many techniques available. The only materials that are suitable for carburising are low carbon steels as the process involves diffusing carbon atoms into the surface of the material. The higher the carbon gradient between the carburising atmosphere and the component the more successful the process is likely to be but it would be possible to write several books on just this subject, as there have been so many variations developed. The most basic technique would be to pack steel into a container of graphite and to heat to 880 Celsius for a period of time depending on the depth of hardness needed. The component would then be quenched to produce a local martensite layer and then tempered to give a suitable surface hardness. (This is also a gross over simplification as depending on the duty required there are many more sophisticated variations used.) One problem with carburising is that the high temperatures used and the subsequent quenching can cause distortion and it is usual to grind parts to finished dimensions after heat treatment. For a gear wheel this is a perfect treatment as the wear regions (gear flanks) can be hardened to about 650Hv whilst leaving the body of the gear as relatively tough and fatigue resistant. The process is obviously good as modern gearbox life is excellent. More recently new plasma techniques have been developed along with Molybdenum spray technology but the fundamental principles are still valid.

Carburising would not be suitable for automobile crankshafts as distortion would tend to be too high and the surface produced may be rather too brittle.


This process is normally applied to mild, plain carbon or very low alloy steels. It is normally carried out in a salt bath. Carbon and nitrogen are diffused into together into the surface of a steel at about 850 Celsius. This can cause distortion unless great care is taken to support the component during heating. Like carburising this process produces a significant increase in surface hardness but the case depths that are usually produced are limited to 0.5mm. As the core materials used in this process are relatively weak there is little point in developing very deep hardened layers and care needs to be taken to ensure that no crushing of the hard layer can take place or fatigue resistance can be damaged. There is a small effect on overall fatigue resistance produced by this process which may be worthwhile for moderately tuned engines.. Nitrogen has a similar effect on hardenability to carbon and a layer of metallic nitrides are produced in addition to the carbon and slightly better properties can be obtained. The slightly lower temperatures used also reduce distortion compared to carburising. Carbonitriding is normally used to upgrade the performance of relatively cheap materials for mass produced components.


Nitriding is most effective when applied to a to a limited range of alloy steels that contain a selection of elements that form stable nitrides. e.g. Aluminium, chromium, molybdenum, vanadium and tungsten. The nitrides produced are dispersed very evenly throughout the surface of the material Nitriding is only carried out on steels that have already been hardened and tempered prior to surface treatment. This procedure produces a core that is very strong compared to the other materials that have been considered. EN40B would typically be heat-treated to a ‘T’ condition (750M Pa tensile strength) before treatment. After this initial heat treatment the component would be ‘stabilised’ at about 580 Celsius to allow any distortion to occur. The component would then be ground to its final size. Nitriding takes place at about 540 Celsius and distortion is normally considered to be non-existent. There is an expansion of the surface caused by the development of the nitride layer and this creates large compressive stresses. These compressive stresses create a very significant increase in surface hardness (potentially 1000Hv) and also cause a significant improvement in fatigue strength. (It is obvious to develop a crack tensile forces must be present and if residual surface compressive stresses are generated then the superimposed stress caused by operation must first overcome the compressive stress before a tensile component can cause damage) Nitriding also reduces the coefficient of friction of the surface of the treated material. (Tractors have ‘frictionless sleeves in the power take off systems and these are nitrided). The depth of case produced by nitriding is relatively limited but as the core material is of high strength this does not cause any practical problems.

I believe that nitriding is the best process to consider for competition crankshafts and EN40B is relatively easy to forge or machine prior to the initial hardening and tempering operation. (NB Many people promote billet turned cranks as being superior to forged cranks. I believe that a forged component will always give better fatigue performance but tooling costs for suitable dies would make limited batch production prohibitively expensive so the industry justifies billet turned as "better". I wish they would be honest and say affordable !)

There are two important points to note:
  1. Never nitride a component that is not manufactured from a nitriding steel. Iron nitrides are fairly soft and spall off the surface of the component and can form a wonderful grinding media when mixed with oil.
  2. Never nitride a component that has not been hardened and tempered as the difference in strength between the core and case can cause surface cracking which would be catastrophic from a fatigue point of view.
I also noted on the [www.se7ens.net se7ens list]? a comment that EN40B or ‘better’ would be suitable. Care is needed when selecting materials. EN41B may be considered better as the surface will be harder than EN40B treated in the same manner BUT the surface of EN41B becomes so hard that it will not tolerate any point loads or bending without cracking. It should only be used for rotor parts such as used in CAV DPA fuel pumps. In order to choose a material you need to consider many factors. In general EN40B has been entirely satisfactory for cranks for many years.

Ferritic Nitrocarburising (Tufftriding)

It is obviously quite easy to confuse nitrocarburising and carbonitriding but they are very different processes. Ferritic nitrocarburising can be applied to almost all ferrous materials. It is essentially a process for upgrading mild and low alloy steels. It can have other benefits associated with friction and galling but these normally only apply to tool steels. In principle this process introduces carbon, nitrogen with traces of oxygen and sulphur into the surface of a steel. With alloy steels it is normal to harden and temper prior to surface treatment to give optimum core properties. The process develops a very thin layer (0.01-0.02mm) of iron carbonitride, Fe2 (CN). The thickness of this layer seems to have little influence on properties. The process temperature is about 570 Celsius , so distortion is limited. As nitrogen also diffuses below the surface layer produced quenching the part will cause a supersaturated solution of nitrogen in the steel (depth will depend on process time) and this will have a very positive effect on fatigue life.

In general this is a valuable process and good results can be obtained with a wide variety of steels and cast irons.

General comments about cranks.

The original 'cast iron' crankshafts produced by Ford for the Anglia 1200 and 1340 Classic were extremely poor. Tuftriding these parts slightly improved wear resistance but the basic fatigue properties of the parent material were very poor and if high revs were envisaged it wasn't worth Tufftriding as the crank would still break. The original steel cranks used in Kent engines were EN40B and were nitrided correctly. These cranks are excellent and were originally forged.

Currently EN40B Ford cranks seem to be billet turned and then nitrided but I am not 100% confident that all of the stages required to produce a satisfactory crank have been carried out and at £1000 each I would need to be confident.

The EN19 Steel Cranks that are available may not be a bad compromise. EN19 hardens quite well to a 'T' condition so the core should be about the same strength as EN40B and Tufftriding should help it to perform quite well.

It is interesting that once a crank has been ground any benefit of the hard surface giving good wear resistance will be lost. The fatigue improvement that takes place in the journal webs will also be lost.

Having a component re-nitrided is quite difficult as a 'white layer' is produced during nitriding and this MUST be removed before use. It is normal to remove this white layer by grinding and an allowance must be left on the component before surface treatment. It could be done but it would need thinking about carefully as you may need to partially re-grind, nitride, then finish grind. As grinding not only removes the 'white layer' but also reduces surface hardness allowances need to be worked out carefully.

Nitrocarburising also produces a 'white layer' but this MUST NOT be removed as it is the hard layer that is wanted. It is fairly easy to re-Tufftride a crank as the distortion and growth which occur are small and well within the tolerance of a crank/bearing clearance to tolerate. The crank journals should be lightly buffed to remove the bloom that develops during heat treatment.

Early Mini Cranks were manufactured from EN16T steel which is medium carbon with almost no alloy additions. They cannot be successfully nitrided but they can be Tufftrided. Compared to an EN40B nitrided Cooper S crank they are not very good and this gives some idea of the improvements that can be obtained.

The latest generation Metro Turbo crank was manufactured from an SG Iron Casting. This is a relatively new material developed at BCIRA and has been known commercially as Duracast. The Metro Turbo crank has a reputation as being stronger than a Cooper S crank and I think that if this material were Tufftrided it would be excellent. Ford USA did produce some Duracast 711M cranks for Formula Ford but I don't think they have any stock left. They also had the extra counteweights rather like the EN40B Lotus Cortina crank that was homologated for Group II (Not standard)

I have seen some of this material tensile tested at Darcast in Smethwick (Volvo crank) and its tensile strength was about 1200M Pa with excellent ductility. I have not seen fatigue properties for this material but I do intend to carry out a literature survey as I I imagine it will be extremely good. You can certainly Tufftride this material to give excellent wear reistance.

Sorry to be long winded but this is a relatively simple overview. I do have a great deal more specific information and I haven't described the different process stages or types in detail (eg Boost Diffusion etc, etc). I can if any one wants.

One more thought: What about con rods. Short of buying steel rods how about Shot Peening to improve fatigue life. This is an excellent process and is used all over gas turbine engines to enhance performance.

The best company to use is Metal Improvements in the South of England as they only use the grade of Austenitic Stainless Steel Shot recommended by Rolls Royce and they understand the basic science.

If you use Guyson shot peening shot you could reduce fatigue life as the shape of the shot is not well controlled and it tends to break in use, leaving jagged edges.

I believe that the increase in fatigue life produce by careful shot peening is well worthwhile.
Chris Flavell

Frequently Asked Questions | Title Index | Recent Changes | User Preferences | Sand Box | Help