DE LOS EFECTOS DE LA PATRIA POTESTAD RESPECTO DE LOS BIENES DEL HIJO

During experimentation it was found that sustained generation of low adhesion was not currently possible (see Figures 10 and 11).

To investigate the possible reasons for these results a model, termed the ‘Adhesion Model’, was developed to relate amounts of water and iron oxide to the adhesion level. The adhesion value is a function of the following key parameters:

• Flow properties (yield stress) of the iron oxide/water mixture • Amount of iron oxide/water mixture in the contact

• Surface roughness

• Friction value in the solid-solid contact (boundary lubrication of the asperity contacts)

• Asperity contact stiffness

The idea of the model is based on the assumption that the overall normal load is partially carried by asperity contacts (solid-solid contact) and partially

carried by the iron oxide/water mixture. In the model the surface roughness is approximated by a zigzag contour (height 2Ra) as shown in Figure 12. The

area of asperity contacts is geometrically approximated by the intersection of the zigzag contour with a plane.

Approximation by

zigzag surface for modelling

2 Ra

h

Surface area of solid-solid contact

Surface area of Fe2O3+ water mixture

Mean roughness

"Typical" schematic wheel/rail surface

Figure  12:   Schematics  of  a  contact  between  rough  surfaces  with  an  additional  layer  of   iron  oxides  and  water  (top);  Approximation  of  the  contact  situation  by  a  zigzag  

contour  for  modelling  (bottom).  

In this model the yield strength τ of the iron oxide/water mixture plays a key role because it determines the separation of the surfaces and thus the extent of solid-solid (asperity) contact. To determine the separation of the surfaces as a function of the yield stress of the iron oxide/water mixtures, the squeeze flow theory has been applied. It is shown in (Covey, 1981) that the static separation h of circular plates in a parallel-plate plastometer (see Figure 13) is a function of the yield stress τ, the radius of the plates a and the applied

normal load N:  

ℎ =  2𝜋𝑎!𝜏

3𝑁 (4)

Figure  13:   Investigation  of  the  flow  properties  of  pastes  by  parallel-­‐plate  plastometry.  

Using the squeeze flow theory requires the assumption that the size of the iron oxide particles is considerably smaller than separation of the surfaces h so that the material between the plates can be regarded as homogeneous. Application of squeeze flow theory is not restricted to iron oxide/water

mixtures, however substances other than iron oxide and water have not been investigated in this project.

Investigations on the yield stress τof iron oxide/water mixtures based on this theory were carried out by Beagley (1976). A significant shear yield stress of the iron oxide/water mixture was only observed for high iron oxide fraction in these experiments (see Figure 14).

2a

N

N

h

Iron oxide + Water mixture

Figure  14:   Shear  yield  stress  a  function  of  iron  oxide  fraction,  from  (Beagley,  1976).  

Figure 15 and Figure 16 show typical results of the Adhesion Model for rough and smooth surfaces respectively. In both cases the boundary conditions were chosen according to the performed HPT tests.

Figure  15:   Typical  Adhesion  Model  results  for  rough  surfaces;  Variation  of  layer   thickness  d;  Boundary  conditions  according  to  HPT  testing.  

Figure  16:   Typical  Adhesion  Model  results  for  smooth  surfaces;  Variation  of  layer   thickness  d;  Boundary  conditions  according  to  HPT  testing.  

Although the model is quite simple and the model parameters need further calibration and validation, the qualitative behaviour of the model demonstrates a feasible mechanism for real world behaviour. According to the model, there is only a narrow range of conditions causing low adhesion. Low adhesion is observed when the iron oxide/water mixture is able to separate the contact surfaces, but is unable to transmit significant tangential stresses. It has to be mentioned, that the model does not take the reduction of layer thickness due to relative motion of the surfaces in to account. Relative motion of the surface (due to creep) is thought to squeeze material out of the contact so that the extent of solid-solid (asperity) contact increases, which should increase the adhesion level.

6.6. Discussion

HPT tests have been shown to useful in assessing the effect of third body materials in the wheel/rail interface when testing in a laboratory. Tests under low amounts of water showed a reduction in adhesion over fully flooded interface. Addition of oxide/water mixtures also led to a reduction in friction. Dry friction levels were as would be expected of steel-on-steel contact and are inline with those previously measured using alternative tribometers (Gallardo- Hernandez, 2008). Under dry conditions there is a reduction in the maximum coefficient of friction for a run-in specimen pair. This reduction is evident as a large shift down when compared to a fresh specimen creep curve in similar conditions and will be related to material work hardening and surface roughening.

The run-in of the surfaces in contact also has the effect of increasing the average roughness of surfaces from initial value (Ra = 0.5 µm) in some cases by up to a factor of 20. Where contaminants were included the increase in the roughness was reduced overall, but there were also localised areas of

roughening. This roughness increase is a draw back to promoting BL between the specimens.

Several issues arose during tests including finding an effective method of application of water and iron oxide. Difficulties arose in the even distribution of the applied third body layer on the rail sample. It was difficult to effectively mix the oxide at water when using high mass percentages. The increase in oxide fraction leads to a rapid increase in the viscosity of mixture that is formed. At low mass percentages the mixture is a fluid, whilst at high mass percentages the mixture produces a clay-like substance.

Stick-slip behaviour is heavily present through a number of tests. Counter measures were trialed to promote a smooth and consistent rotation of the bottom specimen, but have not proved effective. Therefore, it is thought that overriding influence on stick slip is the contact conditions themselves. It is thought that the stick-slip that is prevalent in tests with contamination is possibly a function of an uneven third body layer. The uneven layer produces areas of metal-to-metal contact between specimens that dominates the adhesion levels seen (see Figure 17).

The developed “Adhesion Model” is able to relate the flow properties of iron oxide/water mixtures and surface roughness to the adhesion level under quasi-static conditions in order to study the interplay of iron oxide, water and surface roughness in a qualitative way. There is only a small envelope of conditions that leads to low adhesion: a certain amount of oxides depending on the surface roughness, together with the correct low amount of water needs to be present on wheel/rail surfaces. The model results suggest that with a limited amount of iron oxide/water mixture on the surface an increase in surface roughness can effectively prevent the development of low adhesion conditions. The results of the “Adhesion Model” with respect to the influence of water are qualitatively in accordance with results from High Pressure Torsion experiments where minimum adhesion was found for low amounts of water under quasi-static conditions.

Differences in the absolute values might be explained by the difference in contact conditions between HPT testing and a wheel/rail contact. In HPT testing the surface is cleaned prior to testing, whilst a wheel/rail contact will never have a fully clean and dry contact. The surface stress distribution between a wheel/rail rolling contact (elliptic pressure distribution) cannot be simulated fully by a HPT test (constant pressure distribution). It is these differences in the stress conditions that might contribute to the observed differences in adhesion. However, HPT tests are extremely useful to investigate the behaviour of different third body layers, the influence of different surface conditions, as well as the reduction of the initial gradient of the creep curve.  (Ultra) low adhesion conditions have not yet been found in HPT testing, however, the absolute value of maximum adhesion under dry conditions is higher in HPT testing than those measured on the tram wheel rig (Voltr, 2014).

6.7. Conclusions

The main outcome of the investigations was that significant reduction of friction (over dry conditions) was observed when applying low amounts of water. This reduction is much higher compared to flooded conditions. An adhesion model was developed to get a better understanding of the interplay between iron oxides/wear debris, water and surface roughness under quasi- static conditions.

Ultra-low friction was not achieved in HPT tests, however. This was a result of the difficulties in applying and distributing third body layers and their inevitable evolution during a test.

An Adhesion Model has been developed that can estimate adhesion levels in the presence of different water and iron oxide mixtures The model is in

accordance with the experience that low adhesion is predominately observed with low amounts of water and explains the difficulties with accurately testing these mixtures (As seen in the HPT test results).

Tribological testing guided the identification of parameters key to the low adhesion mechanisms such as roughness and amount of water. Furthermore, the general characteristic of the adhesion level as a function of the amount of water predicted by the Adhesion Model was confirmed under rolling contact

conditions by the experiments on a full-scale test rig where it has been used to support the specification of these tests (Buckley-Johnstone, 2016).

Despite the simplifications and the limitations of the model, conclusions with regard to the interplay of the amount of water, the amount of iron oxide and the surface roughness with respect to adhesion conditions may be drawn for the design of future HPT experiments and wheel/rail contact conditions in the praxis. According to the results a reduction of adhesion due to the presence of iron oxide and low amounts of water in the contact is feasible in quasi-static conditions, however this reduction of adhesion is limited to a narrow range of conditions