Historical Issues and Solutions
Running component wear life and failure are primary cost drivers in the rotating equipment industry.
Great effort has been expended to provide dry running capability, reduce friction and wear to extend life, MTBM and MTBF. Most of the improvements in component design have come in the form of material advancements. The earliest forms of low friction materials were carbons and soft metals, but they had no dry running capability. Harder materials have run the gamut from metals to tungsten carbide to ceramics.
This evolution has essentially stopped at silicon carbides. After 25 years of maturity, they are now the standard in many heavy duty applications of mechanical seals, mag drive bearings and shafts. While these materials are robust when lubricated, they are extremely unforgiving when run dry or in other upset conditions. With 40 percent of mechanical seal failures attributable to operational failures associated with lubrication (Christopher Little, Pumps & Systems , May 2008), significant value can be gained if running faces can survive dry running conditions.
In recent years, low friction coatings have emerged to improve situations. The problem with coatings is that they are typically a dissimilar material added on top of the original component (substrate). The process results in a number of undesirable effects such as increased dimensions. Other effects that weaken the adhesion between the coating and the substrate include excessive process temperatures or dissimilar material properties (i.e., thermal expansion coefficient) that meet at the boundary between the component and the coating. This presents delamination risk such as spalling or peeling. When this is coupled with the risk of the coating wearing out and causing the surfaces to run against the uncoated component, users have often avoided coatings.
Hope for performance improvement was heightened with the advent of synthetic diamond in the 1990s. For all its promise, however, this technology has not deeply penetrated industrial applications for many reasons. Reasons include the expense and limited size of the reactor needed to produce the coating, batch limitations and slow growth rate of the coating. Temperamental process control is required to ensure that a low friction surface results instead of an abrasive cutting surface. Unavoidable issues result from the fact that diamond is still a coating with all the inherent delamination risk, added dimensions and geometrically induced inconsistencies.
New Solution
Researchers discovered a process whereby carbided material, such as silicon, tungsten or other, could be removed from a carbide surface via chemical reaction. In the same reaction, the carbon, already sintered into the substrate, is transformed into various carbon species including planar graphite and particulate nanocrystalline diamond. The process, called a CDT treatment, is controllable to any desired depth. The resulting surface is low friction, self-lubricating, able to run dry or in inert gas, and does not have an adhesive bond line, like coatings.
Since there is no dissimilar material added and no bond line, there is no delamination risk, added dimension or geometric inconsistencies. The process can be performed on finished components, and put into service with no post-process grinding or lapping operations.
A scanning electron microscope image of a cross-section of the resulting surface is shown in Figure 1. Notice that the new surface crystal structure is intimate with the virgin substrate, with no adhesive bond line to delaminate.
Figure 2 includes transmission electron microscope images that show the material constituents at various depths into the converted surface. This is not a homogeneous region, but rather a matrix that changes characteristics with depth. The outer portions contain mostly graphite and free carbon, but some evidence of nanocrystalline diamond onions can be seen in Figure 2.4. At the interface where the converted region approaches the virgin substrate, Figure 2.3, the nanocrystalline diamond species becomes more prevalent. In Figure 2.2, the nanocrystalline diamond species can be seen mixing with the virgin carbide along with the free carbon and graphite. In this zone, the hard carbide is supporting the nanocrystalline diamond in particulate form, and is intermixed with the free carbon and graphite for lubrication.
Figure 2
Novel friction and wear behavior results from this new surface matrix. It is immune to wet versus dry running conditions, as shown in the pin on disc measurements in Figure 3. The test apparatus started dry, then was submerged in de-ionized water. The extremely low friction coefficient was unchanged.
Figure 3
Other dry running tests are shown in Figure 4. This shows that the surface remains low friction in the presence of ambient air, and is even lower in an atmosphere of dry nitrogen where there is no humidity. This is not common behavior in any running surface. A further benefit is that the wear mechanism of the surface minimizes wear debris, making it suitable for clean processes or pharmaceutical mixers.
Figure 4
This running surface was subjected to accelerated life testing in a series of aggressive mechanical seal tests. The seal was overbalanced in an attempt to wear the seal faces to failure, and run with various pairings in flashing hot water conditions. The PV for this condition was in the range of 205,000 psi-ft/min. Figure 5 shows the results for two of the face pairs; one pair was silicon carbide running against itself, and the other pair was CDT-treated silicon carbide running against itself.
|
SiC mating ring grooved, wear scar is 173 µm |
SiC primary ring grooved |
|
CDT treated mating ring flat, wear scar is 4.4 µm |
CDT treated primary ring flat |
Figure 5
The results were dramatic. After 24 hours in this condition, the silicon carbide faces seriously grooved, while the CDT treated faces remained flat. The wear scar on the mating ring was 173 μm for the silicon carbide, while only 4.4 μm for the CDT-teated faces. Similar results were obtained in other tests that ran for 100 hours. Comparison of the wear scar depth in this example suggest that CDT treated faces would last up to 40 times longer than silicon carbide faces. Another implication would be that the leakage, which
