Waterjetting 18c - Abrasive choices

Picking the right abrasive can make a big difference in the profit that a waterjetting operation makes. But the question, of course, is which abrasive is the best? And, as I have done in the past, I am going to hedge a little in my answer. The reason for this is that there are different factors that control the price of the abrasive – how far has it to be transported, how it was prepared and what it is made of – for example. And while some abrasives generally cut better than others, if the unit is only going to be cutting a narrow range of material, then the abrasive that is best for cutting a wide range of materials (garnet) may not necessarily be the best choice in that particular case.

Figure 1. Different types and sizes of abrasive particles.

And further, just to make life a little more complicated, not all garnet abrasive (or other types of abrasive product) are created to give equal performance. As I have mentioned in a previous post the cutting performance of the abrasive can be rapidly reduced as the particle size falls below 100 micron. Thus, if the particles have not been well graded, so that there is a significant amount of abrasive below this size (even though the vendor tells you that it is a 250 micron size) then the cutting performance will not be as good as it would be if the mix contained no fine fraction that far below the stated particle size.

Figure 2. The effect of particle size on cutting performance

On the other hand if particle sizes are too large, then for a given abrasive feed rate there will be a smaller number of particles hitting the surface, and this can also reduce cutting performance if carried too far.

Figure 3. The effect of increasing particle size on the cutting of cast iron at a constant abrasive feed rate (0.88 lb/min) (after Hashish*)

There are other factors that has also to be considered in selecting the best abrasive, and, while I am going to discuss this below, in most cases it is going to be something that you will have to test in your own shop, comparing the results that you get with the cost of the abrasive (often worked out in dollars per square foot of cut, or similar unit) to decide which gives you – in your particular circumstance – the best performance.

But the testing of different abrasives can be reduced to a manageable range of tests (and we prefer the triangle test that I have described in an earlier piece) if some basic thoughts are born in mind.

Decide what it is that you will be cutting – is it mainly going to be a metal that is going to respond in a ductile way when cut by the abrasive. In which case the abrasive should have enough sharp edges to cut into the material, and then to plow up some of the surface so that as more particles hit the surface, pieces are broken off.

Figure 4. Mechanisms for cutting into a metal or other ductile material.

On the other hand, if the material that you are going to cut is a brittle one, say for example rock or glass, then the material is going to be removed by crack growth. Here relatively spherical particles can be more effective because the energy of the particle is concentrated at the point of impact, and more easily causes cracks to grow. On the other hand relatively flat particles with multiple impact points reduce the pressure under any one and reduce the effectiveness of the particles in getting the cracks to grow as quickly and as long as possible.

Thus, for example, we can compare the effect of using the same amount of steel shot (round) and garnet in cutting granite (a brittle material) and tool steel.

Figure 5. Cuts in granite and tool steel using the same abrasive feed rate, but the cut on the right is with garnet and that on the left is with steel shot. Note that the steel cuts the granite to a deeper distance within the cut, while in the tool steel it bounces off without leaving much impression. (However harder steel grit on softer steels can be an effective choice).

The difference that particle shape makes in cutting ductile materials can be shown when the same AFR is used but in one case the abrasive is broken glass particles and in the other it is glass beads.

Figure 6. Effect of particle shape (broken glass and glass beads) in cutting composite material at the same abrasive feed rate (after Faber***)

Further some particles, for example steel shot or grit, can be recycled through the system a number of times without much degradation, so that if they can be simply collected below the cut, they can be profitably reclaimed, providing that the particles below 100 micron are separated out of the flow after each cycle.

Figure 6. Effect of recycling abrasive through a slurry abrasive jet system on depth of cut achieved (after Kiyoshige et al**)

Others however are more friable, this is perhaps more true of alluvial garnet, which often has a much higher density of internal cracks than mined garnet, and so is more liable to fracture either in the mixing chamber or on first contact with the target, so that the amount that remains above 100 microns is substantially smaller than that with the mined alternative.

Yet cost is a factor, and so it is best, for your material to test the different alternatives available, before making a final decision.

* Hashish, M., "Abrasive Jets," Section 4, in Fluid Jet Technology, Fundamentals and Applications, Waterjet Technology Association, St. Louis, MO, 1991.
**Kiyoshige, M., Matsamura ,H., Ikemoto, Y., Okada, T., "A Study of Abrasive Waterjet Cutting using Slurried Abrasives," paper B2, 9th International Symposium on Jet Cutting Technology, Sendai, Japan, October, 1988, pp. 61 - 73.
*** Faber, K., Oweinah, H., "Influence of Process Parameters on Blasting Performance with the Abrasive Jet," paper 25, 10th International Symposium on Jet Cutting Technology, Amsterdam, October, 1990, pp. 365 - 384.

Waterjetting 18b - Abrasive Effects

Abrasive particles were first fed into jet streams as a way of helping to clean surfaces. In many cases the intent was to remove a surface layer of paint/rust so that the underlying surface would be clean and able to accept a new coat of paint or other protective coating.

However, as with many applications of waterjets and abrasive waterjets, putting too much power into a single jet stream could be counter-productive. As an example consider the case where an air-powered abrasive jet (sand-blaster) is used to clean paint from a surface so that it can be recovered. There are several concerns with the use of the abrasive that are not always immediately obvious to the operator. In an earlier post I wrote about the problems of surface deformation. With abrasive particles in the jet stream the operator often works to achieve a smooth, clean surface on the work-piece. The problem that this causes is that the abrasive particles bend over small protrusions on that surface, so that contaminant is trapped under the bent metal and accelerates corrosion after treatment. As a result an abrasive cleaned surface can need repainting after 2 years, whereas with proper cleaning the intervals can be stretched to 5 years. This latter period can be achieved with the use of high-pressure water as a cleaning tool, since this is able to wash into the small crevices in the surface and wash the corrosion products out of the surface, and leaves it clean.

The problem that this creates, in turn, is that it removes all the protective coatings and if the surface is a metal this can lead to flash rusting of the surface and other problems if the surrounding air is corrosive and surface treatment is not carried out fast enough. (In some chemical plants the period available for this coating can be less than an hour). In other cases, such as for example when Linda Merk-Gould cleaned the Statue of Freedom from atop the Capitol Building in Washington, other coatings are required to return the surface to the desired color and texture after cleaning.

Figure 1. Looking at cleaned test panels on the Statue of Freedom with Linda Merk-Gould, during the cleaning.

Further, in that particular case, the antique nature of the metal surface meant that the pressure of the jet had to be precisely controlled in a narrow range where it would be sufficient to remove the corrosion, and yet not sufficient to eat into the weakened metal surface. (That pressure range was IIRC about 3,000 psi).

The second problem that can arise with the use of abrasive on a surface occurs if small particles of that abrasive become lodged in the surface during the cleaning. While in most cases the small individual sizes of the particles have little effect on subsequent performance there are cases where this can be a problem. One is where the surface will be enameled after cleaning, and if steel is used, rather than an iron grit, in the final cleaning any residual steel on the surface can mar the final coating used in the process. Similarly in the advanced welding of structures any garnet or similar abrasive that is left in the surface an affect the integrity of the weld that follows. In both cases, as mentioned earlier a final wash with high pressure water without abrasive can be effective in cleaning the surface to the level needed.

This brings up the topic of the nature of the abrasive itself. In the past sand bas been the most commonly used abrasive, although in more specialized applications small-particle slag, walnut shells and other specialized abrasive is used. This is often the case where (as with the walnut shells) there is a need to minimize the damage to the surface being cleaned – and historically they have been used in cleaning bronze monuments, as a way of minimizing the loss in detail that occurs with harder abrasive, although as the experience in testing for the Statue of Freedom showed, full detail could be retained with the high-pressure water clean, while abrasive would blur the finer detail. Walnut (and other nut shells and parts) were chosen because of their relatively soft nature in contrast with the underlying material beneath the coating being cleaned. The use of softer abrasive, and in some cases soluble abrasive, helps preserve the underlying substrate and can lower clean-up costs, although the abrasive itself might be more expensive.

There is, however, a difference in approach when the surface will be left as cleaned when it is put back into service, and that where it is to be painted or similarly protected before being used again. The need is for the surface to be toothed, or notched, so that the overlying coating can attach to the surface and become harder to remove. (In passing this is why high-pressure waterjets are effective in concrete repair, since they only partially expose the aggregate and the new pour can attach to this rough surface, giving shear as well as compressive and tensile strength. Where the surface has been ground with diamond or carbide wheels the interface is smooth and there is a much lower adhesion between the layers, and the repairs, as a result, don’t last as long).

Figure 2. Microphotograph of a steel surface on the edge of an abrasive jet cleaning path, showing the surface deformation due to individual particles.

In these cases the shape of the particle, the size of the particle and its relative hardness all play a part in control of the final result of the treatment. However, the velocity at which the particle hits the surface is also a factor, controlled by the pressure of the delivery system. Too much pressure will, as mentioned above, bend over the surface protrusions so that the surface becomes smoother and more polished. While this has an aesthetic appeal at the time of treatment the rougher, greyer surface with the protrusions left in place gives the better grip. As Plaster noted in a table presented in his book on Blast Cleaning and Allied Processes, increasing air pressure above a certain value is counter productive.

Figure 3. Change in bond strength as the air pressure during sand-blasting is increased. (Plaster ibid)

I will discuss different aspects of abrasive choice in the posts that follow.