Monday, March 23, 2015

Waterjetting 31c - Jet interactions and traverse speed

The last post included the idea that, with a jet penetrating to most of its final depth in the first 1/100th of a second, that the best traverse speed that could be used would end up around 50 ft per minute. Actually this is a little low as an estimate, and the reason for this goes back to a picture from another post. On the other hand the speed may be reduced a little, as a way of getting more material out of the target faster.


Figure 1. Jets penetrating through a bed of glass beads. The original image has been duplicated and the two shown together. The darker green is where the water has penetrated between the glass beads.

In the picture shown not only is the water penetrating and cutting down into the bed, it is also, to a small degree, penetrating into the side walls of the hole, and this becomes more evident as the jet reaches toward terminal depth, and the jet coming back starts to interact with the jet entering the hole, giving the bulge in the penetration, as the water starts to reach out further into the material.

While it obviously varies with the material, the side penetration of the jet into the material around, and ahead of the jet means that the material is, to a degree, pre-weakened by that penetration, and becomes easier to cut, so that the best speed for cutting can be slightly increased from the number derived last time.

There is another benefit to this, which the Chinese developed in one of their mining machines, although it also has application in milling and other removal techniques.

The high optimal speed of the jet was developed by oscillating a waterjet nozzle vertically, as the head was traversed, more slowly along the horizontal. Because of the benefits of the oscillation the head cut a wide path through the material to about the same depth as it did when the jet was traversed without oscillation. The volume removed however was roughly an order of magnitude greater.


Figure 2. Slots cut into a soft cement. The two slots are roughly equivalent in depth but the lower one had the nozzle oscillate perpendicular to the traverse direction across the block. (see here)

One can, as a result, remove much more material where a plain waterjet (as opposed to one that carries abrasive or cavitates) is moved across the target surface.

When looking for the best combination of oscillation speed, as a function of the horizontal speed in cases such as those shown in Figure 2, it is important to consider how much of the water has penetrated into the side of the cut, as shown in Figure 1.

The softer the material, and the easier that the jet penetrates into the wall of the slot, then the further apart the two passes can be made. But that inter-cut distance varies also with the speed that the primary cut is being made. If the cut is being made at the most efficient speed to remove material, then the speed will be high so that the rebounding jet does not start interfering with the incoming jet, as is shown in the lower parts of figure 1.

Thus the two adjacent passes should be made close enough that the two layers of darker green around the upper parts of the slot depths in figure 1 just barely overlap. On the other hand if the traverse speed is slowed so that the interaction of the incoming and outgoing jets does occur, with the result driving water into the ribs to the extent shown at the hottom of the holes in figure 1, then the two cuts can be spaced further apart, roughly at the distance shown in the figure. This is because any rib of material between the two cuts is now cut on either side and beneath by the penetration of the water, and will be removed with no additional energy being required.

This is particularly handy in materials that have a lot of joints such as, for example, coal or soil. Here the spacing can be increased significantly.


Figure 3. Slot cut into coal by a nozzle with two orifices (upper left) oscillating vertically on the front of the plow shape shown on the right. The slot is roughly 2-inches wide.

However this does not work as well where the target material is stickier, for example with a clay.


Figure 4. Cuts made into wet clay – to help in visualization the cuts are filled with bentonite (white) to contrast the shape and location with the darker clay that was cut through.

In drier clays and shales, where the material responds in a more brittle way, the rib between the two adjacent cuts may shatter and give the larger volume removal rates required.

Where this is not the case, and where the cutting ability of the two jets can be better estimated then two adjacent jets can be aimed so that they meet within the body of the target.


Figure 5. Waterjet cuts in claystone with the two cuts inclined so that they met at the bottom of the wedge shape shown. The wedge is loose, and can be lifted out of the block by hand.

In this way, by arranging a set of jets into a pattern which will be controlled by the shape of the cut to be made, and over what area, a large volume of material can be removed with the use of the least amount of water.

Making the best use of this combination does require some testing to determine how best to angle the jets so that they meet at the right distance into the target and with sufficient force remaining that they will remove the isolated blocks of material between the two (or more) jet paths. In these cases it is best where the jets are traversed over the surface of the target at a slower than optimal speed, since a small buildup in pressure at the bottom of the common slot will make it easier for the jet to dislodge the wedge of material between the two cuts.

As material gets stronger, however, the benefits of that penetration become less, and to make sure that the material between the two cuts is removed, the two cuts should be aimed so that they will intersect within the cutting depths. With vertical slots the two cuts can be brought quite close, with no interaction between them.


Figure 6. Parallel adjacent cuts in sandstone, where the ribs are not removed by the jet between adjacent passes.

1 comment: