**4. Experimentals method**

The substrate material used in this work was AISI 316L type austenitic stainless steel of following chemical compositions (in wt.%): 17.018 Cr, 10.045 Ni, 2.00 Mo, 1.53 Mn, 0.03 C, 0.048 Si, 0.084 P, 0.03 S and balance Fe. This steel was supplied in the form of 2 mm thick hot-rolled plate. Samples of 20 mm x 70 mm size rectangular coupon were cut from the plate. The sample surface was ground on 320, 600, 800, 1000, 1200 grit SiC papers, and then polished using 1 μm Al2O3 pastes to the mirror finish. Before treating, these specimens were cleaned with acetone. The treatments were performed at 450 C for a total duration of 8 hours in an electrical resistance heated fluidized bed furnace having 105 μm particulate alumina as fluidized particles which flow inside the chamber due to the flow of nitriding or

layer formation are consistent with the large compressive stresses in the residual stress

Previous investigation which regards to the influence temperature and time of thermochemical treatments using Fluidized bed shows that nitriding at 400°C for times up to 6 h could not produce a continuous nitrided layer on the substrate surface. When temperature was increased to 450°C, a uniform layer was formed after 6 h nitriding and was not effective for shorter treatments due to only a very thin discontinuous layer formation after 3 h nitriding, whilst after nitriding for 6 h a layer about 13 μm thick was formed, which has a bright appearance and is resistant to the etchant used to reveal the microstructure of the substrate. Increasing the temperature to 500°C resulted in the formation of a relatively thick nitrided layer after 3h and 6 h nitriding. The morphology of the nitrided layers formed at this temperature for longer treatment times is different from that formed at 450°C for 6h. Some dark phases were formed in the layers which is similar to those observed for plasma nitrided product and can be attributed to the decomposition of the S phase and the formation of chromium nitrides, which is believed reduce the corrosion (T. Bell, 1999).

The microhardness measurement on the nitrided subtrate showing a function of processing time for the three different temperatures. It can be concluded that no obvious hardening was achieved after nitriding at 400°C. The hardening effect is also insignificant after nitriding at 450°C for 1 h and 3h, and at 500°C for 1h. This corresponds well with the above metallographic examinations that no effective nitriding was achieved under these

According to these, both structural analysis and hardness measurement indicate that under the fluidized bed nitriding conditions, there exists an incubation time for the initiation of nitriding reactions. Nitriding must be carried out for a duration longer than the incubation time in order to produce an effective nitrided layer. From the experimental results, it is also evident that the incubation time is temperature dependent: increasing nitriding temperature

This incubation time phenomena, which has not been reported for other nitriding processes, such as plasma nitriding, may be related to the nature of the fluidized bed process, where the disruption of the native oxide film on the specimen surface, which is required to effect the nitriding reactions, has to rely on thermal dissociation. The higher the temperature, the

The substrate material used in this work was AISI 316L type austenitic stainless steel of following chemical compositions (in wt.%): 17.018 Cr, 10.045 Ni, 2.00 Mo, 1.53 Mn, 0.03 C, 0.048 Si, 0.084 P, 0.03 S and balance Fe. This steel was supplied in the form of 2 mm thick hot-rolled plate. Samples of 20 mm x 70 mm size rectangular coupon were cut from the plate. The sample surface was ground on 320, 600, 800, 1000, 1200 grit SiC papers, and then polished using 1 μm Al2O3 pastes to the mirror finish. Before treating, these specimens were

hours in an electrical resistance heated fluidized bed furnace having 105 μm particulate alumina as fluidized particles which flow inside the chamber due to the flow of nitriding or

C for a total duration of 8

faster is the dissociation of the oxide film and thus the shorter the incubation time.

cleaned with acetone. The treatments were performed at 450

profiles which were determined by XRD.

conditions.

reduces the incubation time.

**4. Experimentals method** 

carburizing gases. The fluidized bed furnace, which was manufactured by Quality Heat Technologies Pty Ltd has a working chamber of 100mm diameter x 250mm deep with maximum worksize of 70mm diameter x 150mm high. Before charging the samples, the chamber was heated to the treatment temperature of 450oC with the flow of nitrogen gas at 1.05 m3 per hour. Then the samples were charged to the furnace and the treatment gases were introduced and their flow rates were adjusted to meet the required composition, with the total gas flow rate maintained at 0.62 m3 per hour. Table 2. summarizes the process conditions employed in this work. Four different treatments were conducted, including low temperature nitriding, carburizing, hybrid process, and sequential carburizing-nitriding. The hybrid process involved treating the sample in an atmosphere containing both NH3 (for nitriding) and CH4 (for carburizing) for a total duration of 8 h, whilst the sequential process involved treating the sample in the carburizing atmosphere for 4 h and then in the nitriding atmosphere for further 4 h.

Nitriding, carburizing, and hybrid treatments were performed at 450 C in a fluidized bed furnace having particulate alumina as fluidized particles which flow inside the chamber due to the flow of nitriding or carburizing gases.


Table 2. Treatment conditions and their corresponding layer thicknesses.

The specimens were heated by electrical resistance heating. Prior to treating, the specimens were soaked in concentrated HCl (2 M) solution for 15 minutes duration with the purpose to remove the native oxide film that commonly forms on austenitic stainless steel and protects the metal matrix from corrosion. This oxide layer is believed to act as a barrier for diffusional nitrogen transport (Rie, 1996). After thermochemical treatments, the specimens were quenched in water. The treated specimen cross sections were first characterized by metallographic examination. To reveal the microstructure, the polished surface was etched

Low Temperature Thermochemical Treatments

**5.1 Layer morphology and hardness profile** 

Layer Thickness, [μm]

**5. Key results** 

time.

of Austenitic Stainless Steel Without Impairing Its Corrosion Resistance 329

Hardened layers with different morphologies were observed as a result of the various treatment conditions and the thicknesses of the layers produced in different conditions are shown in Table 2. The layer thicknesses are found to be different at different treatments, and their growths against time in Fig. 6. show that layer thickness increases with processing

> Nitriding Carburising Nitrocarburising Hybrid

258

Time, [h]

Micrographs in Fig. 7 show that the morphology of the layer changed with treatment conditions. The two specimens processed under combined treatment conditions, 8(C+N), and 4C–4N produced duplex layers irrespective of whether they were processed simultaneously (Fig. 7b) or sequentially (Fig. 7d). The processed layer thicknesses in Table 2. show that the nitrided specimen, 8N, has a thickness between 3.26 to 8.35 m and the

Furthermore, the nitrided-only 8N specimens have deeper layers than combined processed specimens. The depth of the simultaneously carburized and nitrided specimen, 8(C+N), reaches only 50% that of the nitrided specimen, and the thickness of 4C–4N specimen had only about 45% compared to the nitrided-only 8N specimen after being processed for the

For a similar treatment duration, the Plasma process is reported by Tsujiwaka on 2005 which is produce about 18 m thick layer which is much higher compared to that of the present conventional nitriding treatment in fluidized bed furnace. In plasma process the native oxide layer is removed mostly by bombardment of the plasma gas which is completely absent in conventional fluidized process. This is one of the reasons why convention fluidized bed treatment produced small layer thickness compared to the corresponding plasma nitriding. Previous investigation revealed that nitriding at 450°C became effective after treatment for 6 h where a continuous nitrided layer was produced (Sun, 2006). This is due to the fact that the incubation time phenomena which may be related to the nature of

Fig. 6. Thickness of treated layers measured from micrographs.

carburized specimen, 8C, is in between 1.00 to 3.92 m.

same duration of 8 h due to the half nitriding duration.

in Marble's solution (4 g CuSO4 + 20 ml HCl + 20 ml distilled water). The schematic picture of fluidized bed furnace is shown on Fig. 4.

The specimens were further characterized by microhardness indentation, elemental analysis by FESEM and X-ray diffraction (XRD) analysis using Cu-Kα radiation. Tribological properties were evaluated with a Taber® Linear abraser model 5750 dry slide tribo-tester using an 5-mm diameter AISI 316L collet nut as mate material. The stroke length applied was 25.4 mm under a constant load 600 g. After 3600 cycles of sliding (completed in 60 minutes) having maximum velocity of 79.76 mm/sec, the specimen wear loss was measured by balance to evaluate cumulative weight loss. Microstructures of treated layers were investigated by X-Ray diffraction analysis (XRD) using Cu-Ka (40 kV, 150 mA) and Field Emission Scanning Electron Microscopy (FESEM). The electrochemical corrosion behaviour of the as-treated surfaces was evaluated by measuring the anodic and cathodic polarisation curves in aerated 3.0 % NaCl solution at a scan rate of 1 mV/min. The tests were conducted at room temperature by using a three electrode potentiostat with a computer data logging, requisition and analysis system. Potentials were measured with reference to the standard calomel electrode (SCE).

Corrosion tests were performed electrochemically at room temperature in a flat cell with 3.0% NaCl in distilled water. The flat cell, as schematically shown in Fig. 5, was a threeelectrode set-up consisting of a saturated calomel reference electrode (SCE), a platinum auxiliary electrode and a working electrode (sample). Sample to be tested was placed against a Teflon ring at one end of the flat cell, leaving a theoretical circle area of 67.5 mm2 on the sample surface in contact with the testing solution through a round hole in the Teflon ring. Test control, data logging and data processing were achieved by a ''Sequencer'' computer software. The scanning potential was in the range of -0.5 to + 1.4 V, and the scan rate was 1 mV/s. From the polarization curves, the average values of the corrosion potential (Ecorr), the corrosion current density (Icorr) and the polarisation resistance (LPR) were calculated.

Fig. 5. Schematic diagram of the flat cell used for polarization corrosion test (Li & Bell, 2004).
