Waterjettting 23a - Injecting abrasive into a waterjet

In an earlier part of this series I wrote about the introduction of abrasive into waterjets, and the loss in energy that occurs when the abrasive and the air that transports it are accelerated into the waterjet stream in the mixing chamber of a conventional abrasive waterjet nozzle assembly.

Figure 1. Conventional mixing of abrasive into a waterjet cutting stream.

Because air is conventionally used to carry the abrasive into the mixing chamber, and due to the relatively high volumes that are entrained it is often the case, as Tabitz* and others have shown, that the abrasive velocity exiting the jet is reduced as air volume increases.

Figure 1. Simulation of the effect of increasing air volume and abrasive feed rate on the particle velocity issuing from a conventional abrasive waterjet nozzle. (Tabitz et al*)

Using a higher density fluid to carry the particles into the mixing chamber is a self-defeating exercise, since the heavier fluids also will have to be accelerated to the final velocity, so that if a carrier fluid is to be used, then air is a logical choice. But it can make up some 90% of the jet leaving the nozzle, the water comprises roughly 9% of the remainder, so that only 1% of the jet may be abrasive, and this is the component that does the cutting in harder materials.

There should be a different way of approaching this, and in the early 1980’s Mark Fairhurst, at the time a graduate student in the UK, came up with an answer, which was presented at the BHRA Conference held in Durham in 1986. The initial system was relatively simple, but demonstrated the principles of the approach which was initially known as the Direct Injection of Abrasive Jet (or DIAjet for short).

Figure 3. Initial flow circuit from which the DIAjet system evolved. (after Fairhurst-1**)

The concept of the DIAjet circuit is that the abrasive particles are first loaded into a pressure vessel, which is then closed. When the pump is turned on part of the water flow from the pump feeds into this vessel through two control valves. The first is at the top of the tank, while the second was directed to feed at the bottom of the Tank, making it easier to feed abrasive into the underlying ejector, which mixed it with the main water flow from the pump, and thence carried it to the nozzle. This approach has a number of advantages over that of the conventional mixing chamber. The immediately obvious one is that there is no air added to the system, and the energy imparted to the water by the pump is only shared with the abrasive particles, without the system losses that occur where air is added to the mixture.

As a result the abrasive particles acquire a higher percentage of the water energy, and achieve particle velocities that allow cutting at 3,500 psi and 5,000 psi, whereas otherwise with a conventional system the jets would be at pressures ten times this high (although we will get into some of the caveats to that statement as this segment of the series continues).

In the earliest version of the system (and in some stand-alone versions that developed later, as I will discuss later in the series) the abrasive was added by simply unscrewing the lid, adding the abrasive to the tank, and then resealing the lid. Part of the problem that this causes is that, if the feed is not properly controlled abrasive can be caught in the threads of the cap piece, and this will then gall the threads and rapidly wear out the connection.

BHR, who first developed the machine, overcame this problem initially by using a secondary circuit to feed the abrasive into the pressure vessel, and this could be arranged so that there were two pressure vessels (which rapidly transitioned into pressure cylinders modified from other applications) one of which could be charging, while the second was in use. The basic circuit then became:

Figure 4. Schematic flow for the first commercial DIAjet system (after Fairhurst-2***)

It is perhaps illustrative to show one of the modifications to the design that was made in Missouri, where we used a small pressure-washer pump to feed the water to the pressure vessels, while the abrasive storage (the hopper shown in figure 4) was made from the pressure tank used in high-pressure painting applications. Because the lid of that pressure vessel was not threaded it was quite easy to refill, and the two cylinders were operated alternately. The entire system was designed to fit into the bed of a pick-up truck.

Figure 5. A small portable cutting system based on the DIAjet system. The assembly is mounted on a metal platform, and includes a water reservoir so that it is largely self-contained, and simple to use.

This new way of adding the abrasive to the waterjet feed has been developed for a number of different applications, although, because of the problems that arose in operating valves which control flow that contains abrasive, there have been some problems that have persisted in finding circuit designs that can operate on a consistent basis for the steady cutting applications where long cutting times are needed. But this approach has a number of applications where the abrasive need only cut for a relatively short period of time, during which the valves can function effectively, and where the jets can perform a cutting operation that is difficult for other cutting applications to achieve. It is, for example, possible to use a DIAjet type of system (if controlled properly) to cut through a live explosive detonator, without causing the explosive to go off. But I will talk about some of these developments, and some of the other capabilities of the system in later pieces.

*Tabitz, Schmidtt, Parsy Abriak, and Thery “Effect of Air on accceleration process in AWJ entrainment system, 12th ISJCT, Rouen, 1994 p 47 - 58.
** Fairhurst, R.M., Abrasive Water Jet Cutting, MSc Thesis, Cranfield Institute of Technology, January, 1982.
***Fairhurst, R.M., Heron, R.A., and Saunders, D.H., "Diajet" -- A New Abrasive Waterjet Cutting Technique," 8th International Symposium on Jet Cutting Technology, Durham, UK, September, 1986, pp. 395 - 402.

Waterjetting 11a - More thoughts on Abrasive

In the last post I mentioned that the abrasive particles, which are fed into a high-pressure waterjet stream to form the Abrasive WaterJet (AWJ) cutting tool, can be significantly crushed when mixing with the high-pressure waterjet, and before they leave the mixing chamber. Because of this - depending on the application - the choice of abrasive can play a significant role in how well the AWJ performs. I have mentioned a number of times that the Waterjet Lab is located at Missouri University of Science and Technology. That meant (apropos “show me”) that it was an appropriate place to run comparative tests between different abrasives to find which is the best.

It turns out that there is no one single answer to that question, since the abrasive that was the most economical and effective to use in one case does not necessarily give the best results in another. Which brings me to the first point in today’s post. It is relatively easy to get small samples of the different abrasives that might be used in a given job. Setting up a small series of test runs, in which the different abrasives being considered, are fed to the nozzle and use to cut standard cuts into test samples, is a relatively easy way to find out which is the best abrasive for that particular material and cutting path. However it is best not to use only a single test run, we would generally run a series with three different jet pressures and three different abrasive feed rates.

Figure 1. Table showing the change in optimal Abrasive Feed Rate (AFR) on cut depth at different pressures.

By bracketing the range that is likely to have the best concentration of abrasive for each pressure (which is not at the same abrasive feed rate, or AFR) the best result can be found for each different pressure value, and the most economical and effective choice for the task in hand can be quickly found. It is important to include economics in the evaluation, since there have been a number of cases we looked at where the most effective choice for abrasive in terms of giving the fastest clean cut was not that much more effective than the second place abrasive, and that alternative was sufficiently cheaper that it made more sense to use it.

The pricing of abrasive, however, is not something that it is easy to generalize over, since there are a number of different factors that come into play, depending where in the country you are located. As a rough guidance, however, we have found that garnet is a more universal cutting abrasive than most others, with less extraneous “issues”, and while it can be less effective than other selections in some conditions, in general it will cut more materials effectively and economically than its challengers. Further mined garnet, in general, performs better than alluvial garnet since it does not have the degree of damage within the particles that leads them to fragment more easily in the mixing chamber.

There are, however, more factors that just the abrasive type that have to be considered. There include the particle size, and range, and then, as noted in the table, there is the selection of the AFR to match the cutting conditions on the table.

One of the more neglected factors relates to the amount of air that is used to carry the abrasive from the hopper into the mixing chamber. The person who did more to shine a light into this corner of the technology was Tabitz, in France. (Tabitz, Schmidt, Parsy, Abriak, and Thery “Effect of Air on accceleration process in AWJ entrainment system, 12th ISJCT, Rouen, 1994 p 47 - 58.)

Because abrasive can cut into the parts of the flow meter, the equipment that they used included a trap between the hopper and the mixing chamber, where the particles could be collected, while the air passed forward to be measured and enter the mixing chamber.

Figure 2. Apparatus used by Tabitz in measuring the air flow to the cutting head and mixing chamber. (ibid)

The results from the measurement showed that as the jet pressure increased, so for that particular nozzle design, did the amount of air that was being drawn into the chamber – although you may note that it begins to reach a constant volume as the pressure approaches 280 MPa (40,000 psi).

Figure 3. Effect of increased jet pressure on the amount of air drawn into the nozzle, as a percentage of the total volume of the resulting jet. (Tabitz et al)

The problem that this relatively large volume of air presents is that it has to be accelerated at the same time as the energy in the jet is being transferred to the abrasive particles. The larger the amount of air in the mix, then the greater the amount of water energy that has to be diverted into accelerating the air. This leaves less energy available to accelerate the abrasive itself.

Tabitz modeled the result with a simulation in a computer program, which illustrates, for different abrasive feed rates, how the average abrasive particle velocity falls as the amount of air in the mix increases:

Figure 4. Simulated effect of an increase in air flow on the reduction in average abrasive particle velocity (after Tabitz et al).

Placing small instruments in front of abrasive-laden waterjets can lead to a relatively short life for those instruments, and measurements of actual particle velocities, though they have been made by a number of researchers, have not been as comprehensive as the above chart might indicate.

Nevertheless there is some indication that the above curves are accurate in principle, if not totally real. A jet with very little air might accelerate particles to 1,880 ft/sec, for example. However with 70% air in the mix, then the particle velocity might fall to 1,700 ft/sec, and with 95% of the jet made up of air, then the abrasive particle speed may fall to 1,200 ft/sec. Part of the difficulty in assessment is because of the very short time interval in which the abrasive particles are accelerated while in the mixing chamber. Because the rate of acceleration of the particles is inversely related to their size. Smaller particles are accelerated faster. And this is the counter to the point that was made in the last post about smaller particles cutter less efficiently than larger ones.

Part of the reason for this is that the smaller particles are decelerated faster in air than larger particles. The results of this in terms of cutting power is one of the areas that still requires more research. If, for example, smaller particles are used in an application (for example to achieve a finer detail in the surface cutting) then the effective range of the jet can become smaller than with larger particles. There are some caveats to that statement, and I will go into some of that explanation in the next post.