Read Frame Fabrication text version
How To Build and Operate Sluice Boxes
Part II Riffle Designs
By Gary Weishaupt
Sluice Box Building Part II - Riffle Design
What we prospectors call `riffles' are things that were invented long ago by good old Mother Nature when she decided to make river and creek beds all rough and broken up thereby creating inconsistencies that broke up the smooth laminar flow of water so that suspended solid materials became stratified and tended to settle, according to their weight, into both small and large irregularities at the bottom of these watercourses. If you take a look at any typical stream you'll see the effects of these natural riffles all around you in the form of what we call `bars'. You'll observe that the material is lain down in distinct layers consisting of heavy materials, lighter materials and then even lighter materials one on top of another. Bars are created where the stream has a preponderance of natural irregularities that cause the materials to drop out of the liquid suspension. In natural watercourse these irregularities can be both horizontal and vertical. Horizontal riffles are created where streams or rivers are diverted laterally at bends and twists while vertical riffles are created at changes in the elevation or slope of the watercourse. A big boulder that rolls down into a creek creates a really great natural riffle that is both vertical and horizontal. When we build a set of riffles for a sluice box we're just trying to duplicate this natural condition of a streambed but in a much more sophisticated and controllable fashion. The evolution of riffles with respect to sluices goes so far back into time that there are no written documents that describe when, where or how some guy first got the bright idea of putting a stick or some rocks perpendicular to a water flow to capture the heavy materials that were suspended in the laminar layers of the watercourse. It most probably came about accidentally based on the observation of naturally occurring riffles like boulders in the middle of a creek. We do know that riffles in the early sluices we're all familiar with consisted of relatively small sections of native timber, sometimes quartered and squared but often just in the natural `rounds' of a tree branch lodged in between the sides of the sluice flume. Almost anything that breaks up the laminar flow of water that carries suspended solid material will create irregularities that cause that suspended material to drop out of suspension. This is how a sluice functions and the process is both extremely crude and yet extremely sophisticated at the same time. Technically speaking, sluice boxes are `gravity concentration' devices, as are regular old gold pans. They operate by taking advantage of the fact that gold is 19 heavier than water and about 6 times heavier than almost all other minerals, rocks, gravels and sands. What is surprising, considering our modern day level of technology is that even a crude wood sluice box with wooden riffles will capture just as much gold as the latest and
Sluice Building greatest hi-tech contraptions we've come up with so far, that is, with respect to gold of a specific size, shape and weight. This last statement is kind of a loaded one so we need to look at what is being exactly stated in more detail. Back in the late 1800's when mining engineers began to take a serious look at sluices they discovered that one could recover what was called `spherical' gold using almost any type of sluice riffle. The term `spherical' here applied to any piece of gold that had any type of `body' to it even if it was elliptical, spherical or irregular in shape just so long as it was not significantly `flattened'. The reader can do their own research here if they want to but most experts, even modern experts, agree that most pieces of gold that have a significant `belly' or `body' down to small pieces that are one sixteenth of an inch or more in any dimension can be recovered relatively easily behind almost any type of riffle design. Some experts who have made exacting studies of riffle designs have lowered this capture threshold down to pieces that are only 1 mm (.03937") in any dimension. Basically this means that almost any piece of gold having a dimensional shape that is approximately the size of a typical pinhead is recoverable in almost any type of sluice equipped with almost any type of riffles. Of course anything larger than this is almost a no-brainer with respect to capture as long as the sluice is set up properly. The minimum sized nugget just described dimensionally is what I call a `micro-nugget' or a `mini-nugget' and they can be incredibly small but they are very different from we call `flakes', `flecks' or `flour gold'. These micro-nuggets can usually be picked up with a pair of tweezers fairly easily as opposed to `flakes' even though these flakes sometimes appear to be much larger in size. The big difference is the cross-sectional dimension of the pieces. The micro-nuggets are indeed just very miniature sized `hunks' of gold while `flakes' are very often flattened to the point of being as thin as gold leaf. There is a huge difference in how these two different types of gold behave in a sluice box. Of course there are also thousands of pieces of this small gold that are in various states between being round, elliptical or flat so there are transitional states of behavior as well. Flat pieces of gold, even those looking very large when viewed from above, have hydrodynamic properties and can behave much like an aerodynamic airplane wing. As water flows past a flattened gold flake the flow over the top portion is faster than the flow moving past the bottom surface and this velocity difference creates a low pressure area at the upper surface so that it can literally `fly' or `plane' in the water if the velocity is high enough. This phenonema is called `saltation' and is the primary reason it is so difficult to concentrate very fine flake type gold. In very high water velocities even relatively large flattened micro-nuggets are subject to saltation. Since gold is so dense particles larger than about 1/16 of an inch (mini-nuggets) in thickness will be rolled or tumbled instead of lifted. 2
One of the characteristics of saltated particle flow is that the trajectory of the particle is parabolic and of relatively short length, especially for heavy minerals so the small pieces of gold that are lifted move through the flume in small arching `leaps' instead of being carried away in long arcs in the laminar flow. This is why the use of a capture fabric like Nomad carpet (miners moss) is so effective since the particle on the downward portion of the trajectory has a tendency to bury itself in the fibers. The objective of `modern' sluice boxes and riffle designs is to maximize the recovery of these very small pieces of gold, spherical, flat and anything in between and to collect the larger pieces as well. There is a current myth that the so-called `Hungarian' riffle design is a state of the art piece of technical engineering but in fact there has been very little actual testing and comparison of various riffles under scientific conditions and little data has actually been published. There is no doubt however that the so-called Hungarian design performs well but not necessarily any better than other alternative designs when one considers other factors that affect the overall performance of a sluice box.
One of the most critical factors in sluice operation with respect to improving performance regardless of the riffle design is the classification and preparation of materials. The term classification simply means running raw materials through a mesh sieve of a specific size prior to feeding that sifted material into the sluice. There are several different standards for laboratory sieve mesh sizes and designations including the American Standard, Tyler Standard, British Standard and others but the Tyler Standard seems to have the widest adoption by most mesh manufacturers. Table 1 tabulates some of the more common mesh designations and corresponding dimensions for both the U.S. Sieve Standard and the Tyler equivalent designations.
Table 1 Standard Laboratory Sieve Size Designations
U.S .Sieve Size
4 5 6 7
4 5 6 7
Opening in mm
4.76 4.00 3.36 2.83
Opening in Inches
0.187 0.157 0.132 0.111
8 10 12 14 16 18 20 25 30 35 40 45 50 60 70 80 100 120 140 170 200 8 9 10 12 14 16 20 24 28 32 35 42 48 60 65 80 100 115 150 150 200 2.38 2.00 1.68 1.41 1.19 1.00 0.841 0.707 0.595 0.500 0.420 0.354 0.297 0.025 0.210 0.177 0.149 0.125 0.105 0.088 0.074 0.0937 0.0787 0.0661 0.0555 0.0469 0.0394 0.0331 0.0278 0.0234 0.0197 0.0165 0.0139 0.0177 0.0098 0.0083 0.0070 0.0059 0.0049 0.0041 0.0035 0.0029
Scientific (laboratory) sieves, the types used by soils engineers and geologists, are sized according to the values in table 1 but take note that most of the sieves, normally called `classifiers' sold by mining supply companies are made of relatively cheap industrial metal screening material (woven wire fabrics) and are not manufactured to high dimensional standards. For example most mining supply outfits sell eight separate classifiers numbered from 2 to 100 as tabulated in Table 2.
Table 2 Woven Wire Fabric Classifiers
2 4 8 12 20 30 50 100
Wire Grid Size
0.5" 0.25" 0.125" 0.08" 0.05" 0.03" 0.02" 0.01"
0.437" 0.203" 0.097" 0.060" 0.034" 0.0203' 0.0110" 0.0055"
Sluice Building You can see how these wire fabric sizes are calculated since the designations are based upon a certain number of openings per inch (mesh) as a number 8 just has 8 openings per inch so the wire grid is spaced 1/8 of an inch apart (.125"). Since the wire takes up some of the space the net opening in the mesh is significantly reduced from the nominal mesh designation. For example a number 4 classifier that most people assume will pass 1/4" material will in fact only pass pieces of gravel that are less than 7/32" or just slightly larger than 3/16". Regular old window screen by the way is typically a number 16 mesh having a 1/16" wire grid with 0.0395" net openings. In 1968 the USGS Geologist Harry Tourtelot (17) developed what he called the `Hydraulic Equivalence Factor' that almost overnight changed the way commercial placer operators looked at the entire process of sluice box gravity concentration. Basically this theory was based upon the observation of the layering and horizontal distribution of various materials observed in sluice boxes prior to their being cleaned. Tourtelot realized that this phenomenon was due to the effect of each particles specific gravity in relation to its physical size. The chart of these equivalencies is included as Table 3. Tourtelot was not the first person to observe this phenomena but he was the first one to look at from a scientific perspective. Miners realized almost a century earlier that there was a systematic deposition of materials in sluice boxes and that there relationships between gravel size, gold particle size and water velocity so in these early days a rough set of guidelines were developed that many of us still adhere to even today such as sluice slope, classification of source material and control of water volume and velocity. Tourtelot recognized that there were very specific scientific reasons behind these old generalities and that a person could operate a sluice in such a manner as to maximize or optimize the recovery of extremely fine gold that up until then had normally been washed out in the tailings. (This is one reason old tailings can be such good places to revisit). Briefly stated the Theory of Hydraulic Equivalence holds that a certain volume of water run at a velocity suitable to wash gravel of a specific size through a sluice will reach a point of mechanical energy equilibrium within two to three feet of the head of the box. At that point there becomes a relationship between the size of that gravel and the size of gold particles that is locked into place by the laws of physics regardless of the design of riffles being used. For instance if you have enough water to wash gravels of 1/4" in size thru the length of a sluice then the finest gold you can successfully recover, with any assurance, will be approximately larger than .025" (40 mesh or larger). After Tourtelots findings became public the recovery of gold down to 150-mesh at large commercial placer operations became relatively common thought not necessarily efficient or consistant.
Table 3 (From Tourtelot 1968) The secret is in proper classification of materials in proportion to the smallest average size of gold contained in any gold bearing materials. For all practical purposes most amateur gold miners will never need anything finer than a number 8 classifier and you'll find that about 90% of all classification work done on placer sites, even by large commercial companies, is now done using number 4 mesh 6
Sluice Building (.25" grid). I personally almost never classify anything down below a number 4 mesh but there are advantages to going down to a number 8 or even a number 12 mesh if you're trying to recover very fine gold. The problem is that classification takes time and the finer you attempt to classify material the less time you have to actually process it. In the mining business time really is money and most of you already know how many hours we all put in to our claims and favorite prospecting sites. Studies have shown however that fine gold recovery goes up almost exponentially in proportion to how finely graduated the source material is classified before the gravels are run through the sluice. This is so important that in some spots you literally may be forced to take the extra time to classify down to a number 12 mesh, perhaps finer if necessary to optimize fine gold recovery. Figure 1 below shows some typical river gravels that have been classified thru a number 2 (nominal 1/2") miner's sieve.
As you can see this type of classification is pretty crude and this certainly isn't the type of material most people would want to pan but I still see folks running this coarse mix, sometimes even larger materials through their sluices.
Sluice Building Figure 2 is a snapshot of the same gravel after having been classified down through a number 4 miners' sieve, which is nominally a 1/4-inch mesh.
Now we're getting down to gravel that is far more in proportion to the size of gold that might be found at most sites today. This is the size of material that I try to run whenever possible even when just prospecting a site for the first time. Initially at an unknown site I might pan some unclassified gravels just to get an idea of what's there but once a decision is made to run the sluice box then we'll take the extra time it takes to maximize our chances of finding some worthwhile flakes and fines. Figure 3 shows the same material again but this time classified past a number 8 sieve and to most people that batch has the consistency of course sand or very fine gravel. Beach sand by the way is about a number 12 mesh. This isn't the best photograph as it makes the material appear to be much finer than it actually is which about twice as course as sand you'd use to mix concrete I seldom go down this far unless I'm finding a predominance of fine gold in batches of #4 gravel that are being run in the sluice but if that is the case then going down this next level almost always doubles the recovery of small flecks, flakes and what many call `powder' or `flour' gold. If you get down to this level you really have to have a lot more patience than I have as the work becomes tedious and then you're faced with secondary recovery to separate the flour from the black sand. Whether this is economically realistic depends on how much gold is there. It some cases it's well worth the extra time and effort.
What is really needed, especially for us part-time prospectors is a number 6 mesh classifier but I do not know any manufacturer making one even though this is a fairly commonly available woven wire fabric. The number 6 fabric comes in two different wire gauges (0.035 and 0.047") so you can have 0.132 or 0.120" openings. Either one of these would work out very well for preparing average sluice box input gravels that would be a better size for the riffles found in the typical Keene and similar type sluices. There is also a number 5 mesh fabric and I'm thinking about making a classifier from this fabric as an experiment. The material coming through a number 4 mesh is actually to large to be optimum for most small non-commercial sluices if fine gold makes up the bulk of the native gold bearing gravels. Keep in mind that the foregoing statements apply to initial field classification of material intended for the stream sluice or high-banker and not to a secondary recovery system that you might use to classify your concentrates at home or at the stream bank once the original batch of materials have been collected. For secondary recovery systems it is not uncommon to classify all the way down to a 100 mesh sieve. All of this technicality is well and good up to a point but then we have to think about what we all deal with everyday in the field since there is no way to memorize all of these various mesh and particle sizes. Most of us can relate to things sized in relation to a pinhead or a window screen mesh and these are the two things I want readers to keep in mind as we go forward in this article. Studies have shown that almost any sluice system can successfully capture about 95% of all gold that has a physical dimension of 1mm or larger. That's roughly equivalent to small pieces just slightly larger than 1/32" or just half the size of material that will pass through a typical window screen. This is incredibly 9
Sluice Building small material and I'd be happy just collecting about 50% of this stuff in my sluice boxes but this seems to be the demarcation point most experts use in defining the threshold of what we call `fine' gold. In my personal opinion any gold smaller than this is what I call `flecks' or `flakes' and this brings us around to some definitions of terminology used in the mining industry. By most standards a `flake' of gold is rather large in outer diameter, sometimes up to a quarter of an inch but it is extremely `flat', sometimes thinner than a piece of paper. A `fleck' is similar in thickness but much smaller in diameter, often almost microscopic in size. A `speck' is just as the name implies and just a flat piece of gold almost so small as to go unnoticed by the human eye, like a typical `fly-speck' appearing in the bottom of a pan which is around .002" in any dimension. There is a scientific standard that is often applied to gold to accurately define it precise shape and form and that's called the `Cory Shape factor'. This factor is simply a number derived from the square root of a nuggets length times it's width divided by its depth. A perfectly spherical or cubical piece of gold has a shape factor of 1.0 while a typical flake might have a factor of 0.125. Measuring individual pieces of gold is not something most of have time for so the terminology outline above is more than adequate unless you're going to be involved in some pretty complicated testing. Just for the readers point of reference when talking about meshes it's sometimes easier to be able to relate the numbers back to common household materials. As mentioned, window screen is 16-mesh, beach sand is typically 12-mesh. Ground Coffee is 20-mesh, sugar is 30-mesh and table salt is 40-mesh. Baking flour is 80-mesh and talcum powder is 100-mesh and finer.
Wetting or Washing Most experts agree that `wet' classification is far superior to `dry' classification so if at all possible try to do your initial classification using a water source to thoroughly wash the source gravels as they are passed through the sieves. An additional advantage of adopting this practice is because studies on sluice box and riffle designs have also revealed that fine gold recovery is significantly improved if the material is thoroughly wetted prior to being placed in the sluice. Simply dropping dry material, even if finely classified, into the water stream of a sluice will degrade overall performance greatly but this is a common procedure adopted by many prospectors including myself and I should know better. It's very easy to write or talk about stuff behind a desk or on a prospecting forum but a very different thing usually occurs when any of us are actually in the field trying to work a particular site so there is a fine line between executions of procedures on the computer 10
Sluice Building desktop compared to what actually happens on the riverbank. I think all of us try to strike a workable compromise between what is theoretically perfect and what works in practice.
Material Introduction and Feed Rate Studies on sluice designs have also revealed that the angle or elevation that material is introduced to the sluice has a considerable impact on performance and that the closer to the water flow, hence the bottom of the sluice, gravels are initially placed the better fines are recovered. In fact the ideal situation is to introduce the materials from underneath the water flow but for us amateurs such a feed system is impractical. It has also been found that a steady, slow, continuous introduction of the source material is critical to fine gold recovery in all types and designs of sluices (or high-bankers). Ideally the material needs to be placed in the sluice as low as possible and as slowly and steadily as possible regardless of all other factors and this is something that is extremely hard to do in the field with the gear most of us carry without taking extra care and spending much more time feeding and tending our boxes. The extra effort will pay off in the long run.
Stratification Studies have also shown that the use of a `slick-plate' immediately ahead of the input or `dump' area aids significantly in allowing the source materials to become stratified in the water stream before they enter the riffled concentration area. Proper stratification is a critical component to successful concentration of fines. Many of the small sluice manufacturers have caught on to this fact and are starting to add slick plates to many of their boxes and this design improvement, perhaps more than fancy riffle designs, is the reason these new generation of boxes are better at fine gold recovery.
Water Flow/Velocity/Sluice Slope Every sluice box set up in any particular stream or river being fed a certain size and consistency of material will need to be arranged to control the volume of water entering the box and the velocity of that water flow with respect to the size and type of material it is trying to process. To a large extent the design of the riffle/matting system in the sluice and the size of the material particles being processed determine how much and how fast (or slow) the water flow must be.
Sluice Building Cleaning Every sluice box must be periodically cleaned of processed material otherwise the riffles and matting, if used, will become clogged with material and simply cease to function effectively.
Performance Factors Summation The reason we've outlined these basic performance factors is because they are crucial elements of sluice operation that improve the chances of gold recovery regardless of what type of riffles, screening and matting is being used in any particular box. To summarize then there are six basic factors that need to be considered for optimal sluice operation and these are: 1. 2. 3. 4. 5. 6. Classification Wetting or Washing Material Introduction Stratification Water Flow/Velocity/Slope Cleaning
As mentioned in the introduction the earliest known riffles consisted of native rocks laid down in natural watercourses to trap gold nuggets. Later in time tree branches and limbs were used and the first matt under-lays were used in the form of sheepskins that collected fine gold in the wool, hence the `Golden Fleece' of ancient legends. As time progressed fabricated sluices were invented, often made from terracotta and later from wood planks and riffles became shaped and worked pieces of wood. Riffles were called `cleats' back then and they were generally individually removable to make cleanup easier. The invention of what we now call the `Long Tom', generally considered to be the forerunner of small sluice systems is usually credited to the Miwok Indians of Northern California around 1849. Other historians claim that the device was imported to California in 1850 by a group of miners from Georgia who worked Greenhorn Creek that feeds the Bear River. I tend to believe the story about the Georgians since there is ample evidence to prove that the `Long Tom' and what was then called the `Box Sluice' were both being used back
Sluice Building east during the Georgia gold rush of 1829. Without these production tools the California rush of 49 would not have been very successful. Up until around 1880 almost all sluices used milled wood battens as riffles, usually 1" by 1" in small boxes up to 1" by 2" in larger equipment and then there was a transition period where metal riffles began to appear. John Golding invented an simpler and more economical method of making expanded metal mesh in 1883 and by 1890 we began to see some of the larger commercial sluices fitted with continuous mesh beds and metal riffle systems so even though the sluice boxes themselves were still made of wooden planks they very much resemble the sluices that we use today. Over the past 180 years since the Georgian `Box Sluice' was invented there really hasn't been a tremendous amount of research undertaken on box or riffle design. In fact I literally spent weeks doing research for this short article and only found a dozen or so books and professional papers on the subjects. Some of the material however contained a wealth of information. Most people, including myself, apparently assumed that the various manufacturers who supply sluices do an extensive amount of research on their products but this does not seem to be the actual case. In fact the so-called `Hungarian' riffles that are considered to be the cutting edge of modern riffle design are first described in a mining book written by Arthur Phillips in 1867 but are also described as far back as 1829 in the individual journals of the Appalachian miners. After doing more research I discovered that the phrase `Hungarian Riffles' describes the practice of placing removable transverse battens or cleats along the bottom of the sluice at closely spaced intervals and doesn't have anything to do with the actual shape of those battens. The phrase is meant to describe the method and manner of placing riffles and not the riffles themselves. I also found references to so-called Hungarian style riffles that are designed very much along the lines of the Gould `Gravity Traps' so I guess that almost everything that we consider to be `modern' has been tried at some point in the past. Before we get into the specifics of riffles it is necessary to address the issue of using some type of recovery `matting' below the riffles. Again, this is not a new idea as miners have been using dried grasses and various types of woven vegetable fiber mats and animal skins in the bottom of sluices for centuries. I'm old enough myself to remember when we used the sisal mats from evaporative coolers as an underlayment long before `miners moss' was invented. For fine gold recovery there is no doubt that underlayments improve the performance of any sluice so as we move along in this article the reader should be aware that the use of secondary catchments fabrics of any type under the riffles is typically used even if not specifically indicated or shown in diagrams.
Sluice Building The reader is also reminded that for the sake of simplicity we will refer to the `back' of the sluice box as the end where water and material is introduced and the `front' of the box is where the water exits the flume. Likewise the `back' of a riffle is the face exposed towards the back of the box and the `front' of the riffle is the face or side the faces downstream of the water flow.
Riffles of any type basically serve the function of being small dams and irregularities that interrupt the laminar flow of water containing suspended solids. As the flow of water and material passes over a riffle turbulence is created and the flow at the position of the riffle slows down very slightly, just enough so that the heavier solids in the slurry drop out of suspension and are trapped behind or in front of the riffle. Figure 4 is a cross sectional view of a portion of a hypothetical sluice box and illustrates this basic principal and shows the old traditional style of solid wood batten type riffles.
As can bee seen in this illustration the heavier solids drop out of suspension due to the slowed flow and turbulence created by the batten riffles and in this case become trapped behind the riffles. If no new materials were added to the sluice box the action of the turbulent flow would continue its behavior and begin to dislodge some of the heavy solids and eject them back up and into the laminar flow where they would be washed completely out of the flume. This action is called `scouring'. If you look closely at a sluice in operation you will observe that the material behind the battens appears as if some unseen force is vibrating it. The smaller particles of sand and 14
Sluice Building gravel look like they are constantly jumping up and down and occasionally you will see small particles and even relative large pieces of gravel suddenly picked up and ejected up and over the riffles as they are sucked up into the laminar flow of the water. The illusion of vibration is created by small vortices of whirling water in the turbulent zone created by the interference in the laminar flow by the riffles themselves. Notice in the picture how the flow of water becomes smoother the further away it is from the riffles. This is the laminar zone of smooth flowing water and in order for a sluice to operate properly there must be a layer of heavy non-turbulent flow over the turbulent zone otherwise flow will begin to froth as air bubbles are formed by the turbulence. Fine gold will mix with these air bubbles and be washed straight out of the sluice. This excessive surface turbulence is typically called `boiling' and is to be avoided at all costs. As a general rule of thumb most sluices perform best when there is about 2-inches of water depth above the top edge of the riffles. This depth will vary somewhat based upon the width of the particular sluice box as narrow flumes require less and wider boxes will require more to avoid this surface turbulence.
If you have enough water flow and velocity there is no reason why you can't run a sluice that is perfectly level but it would not be very efficient since such high flows would be necessary to carry away large heavy pieces of gravel that all the smaller gold would be carried away as well. For this reason sluices perform best when they are sloped towards the outlet end so that gravity will assist in the removal of larger materials without relying on high water velocity by itself. The amount of this slope is dependent on the velocity of the incoming water and the size and mass of the materials being processed. For most small sluices the slope will range from as little as ½-inch per foot of sluice length to as much as 2-inches of slope per foot of length. One inch per foot seems to be the preferred setting to use during the initial setup. Sloping a sluice also has a secondary effect and that is to reduce the depth of the heavy laminar water flow that lies on top of the turbulent flow created in the region of the riffles. As the slope increases the upper layer of water moves much faster than the lower turbulent layer so the depth of the laminar flow decreases substantially. If you have excess slope and insufficient volume the water running down the sluice will be entirely turbulent flow that is taking in huge amounts of air and the surface water layer will become froth and foam which will carry away almost all fine gold trapped by the riffles. Figure 5 illustrates a sluice with low water volume and excessive slope so there is almost no laminar water flow above the riffles so the entire bed of the sluice is seen to be in turbulence causing air to become entrapped thereby creating foam at the surface layer. Since gold is hydrophobic (not wetted by water) the small particles will actually attach themselves to the surface of these air bubbles and be pulled up from the bed of the sluice box by the action of `boiling'.
Sluice Building This should not be confused with the normal small wavelets you will see on the upper layer of the laminar flow as it passes by the region of a riffle. There is a fine line between proper surface flow velocity and the point at which it becomes too turbulent and begins to entrap air bubbles. This is not something that can be fully described in words and has to be seen and experienced in the field to be fully appreciated. Boiling is not usually much of a problem with small hand sluices but it can become a critical factor when operating a high-banker, power sluice or suction dredge where the control of water volume and velocity is much harder to regulate. Some experts believe that excessive velocity and boiling accounts for at least 50% of fine losses in typical power equipment today but this is nothing compared to the 90% losses that were typically experienced by the old-time dredge operations, which is why old dredge tailings are so rich in material since it's already been consolidated.
The opposite case can occur where the sluice box has insufficient slope and/or an excessive volume of water flow and the action of the riffles to create eddies and vortices are hindered by the weight of the deep laminar flowing water. This situation is usually seen in sluices in the front portion where the foremost riffles don't appear to have any turbulence at all because the box is sit to deeply in the stream. The weight of the excessively deep water will largely cancel out the effectiveness of small eddies over the riffles. When this begins to happen you will notice that the small wavelets occurring over the region of the riffle will almost completely disappear and the water surface will look almost flat and smooth. In summation, any style or design of riffle must have a proper amount of water volume, specific water velocity and proper slope in order to perform its function effectively. These three factors are far more important for gold recovery in general than any unique design characteristic of the riffle. There are however certain riffle shapes that are better at 16
Sluice Building collecting gold particles of specific sizes but even when they are utilized the sluice must still be set up properly for optimum recovery.
Riffle Evolution and Design
Miners are pretty smart people by nature and it didn't take them to long to figure out that sluice performance could be significantly improved by messing around with different shapes, sizes and spacing of the riffles. For a lot of business people back in the old days this was like inventing and marketing the `better mousetrap' and all kinds of patents were applied for a variety of sluices and riffles but since the operational principals were not `unique' most of these patents were declined but that didn't slow down the promoters back then and they're still at it today trying to claim 99% efficiency if you buy their special products. Somewhere along the line, perhaps as early as 1860, miners began having problems with the tops of their old wooden batten riffles being chewed away by the gravel slurries so they began adding metal plating or straps to the upper surfaces and apparently a few engineers noticed that if these metal plates were accidentally set slightly off center creating a small lip on the downstream side a special type of eddy was created in the water flow that acted like a small backflow. This phenomena we now know is just a low pressure area on the front side of a riffle that permits heavy solids to almost immediately drop out of suspension and accumulate in a small pocket of almost completely still water. Figure 6 illustrates how the old original wood riffles that were worn down by gravel were `rejuvenated' with the metal caps.
Sluice Building The lip created by these caps on the front side (downstream side) aided in creating small vortices or eddies that allowed heavy materials to settle into these low-pressure zones. Of course even the old rectangular wood battens had downstream low-pressure areas but they were not nearly as pronounced as those created by the metal lips. This evolutionary process, accidental as it was, lead to the invention of what can be called the `counter-current' theory in the shape of sluice riffles where both the upstream and downstream faces of the riffle contributed to the collection and concentration of heavy solids. By a misuse of terminology these compound shaped riffles were called `Hungarian' riffles. Word spread fast and this idea caught on rapidly so everybody and their brother almost immediately stopped using the old tradition wood batten riffle and switched to metal riffles in dozens if not hundreds of different configurations. Figure 7 illustrates just a few of these different designs being used by 1867 and many of them are still used today in both large-scale commercial operations and by us small-time prospectors.
What is ironic is that almost all of these `compound shaped' riffles work extremely well if all other factors of the sluice box set up are paid attention to, meaning the box slope, water volume, water velocity, controlled feed rate, use of underlayments and periodic cleaning. As mention earlier the specific shape of the riffle is perhaps the least important 18
Sluice Building factor in the overall performance of the sluice but it is the one factor we have the most control over since it is the single static element in any box. I like to think of it as the box and riffles being the hardware and the proper setup being the software in a simple gold concentration and recovery machine. The earliest metal riffles were like that shown in figure 4-G above which were basically just flat iron bars, usually 1/4"-inches thick by 1 to 2-inches high, which were set either upright, or sloped at an angle as seen in the sketch. From a performance standpoint this riffle style was long thought to be almost ideal and is still, in various guises, used by almost all modern sluice manufacturers. Figure 8 is a snapshot of a typical flat bar riffle system made from 1/8x1" steel strap which closely resembles examples from the late 1800's. The slope of the bars in this example is 30-degree from the vertical but I have seen similar units that have much greater slope sometimes as much as 60-degrees from vertical but most makers seem to prefer a 45-degree slope.
Figure 9 is a picture of another flat bar type riffle system except this time the bar is manufactured from a section of thin 16-gauge bent plate set in the frame with a 45-degree slope.
That portion of the bent plate that lays horizontal and flat against the bottom of the sluice bed above the mesh serves no functional purpose and is only utilized for ease of manufacturing and installation. If a mesh or matt it used it does however help to seal off the bottom of the riffle better than a plain flat bar used by itself. I have used both of these sluices on a regular basis and cannot ascertain that either one has an advantage over the other with respect to collecting fine gold or larger flakes but they do have different characteristics even though they are basically identical in design concept. The 1/8-inch thick bars in the one example produces what I consider to be a `smoother' and slightly larger or longer vortex than the example made from the thin 16-gauge stamped steel which has a very sharp, short and smaller downstream eddy. As stated earlier, this design of riffle is still extremely popular both in large-scale commercial operations and small-scale hobbyist sluices/high-bankers such as The Jobe Yellow-Jacket, Pro-Line, Gold-Buddy, Buckabilly, Honcoop and some Keene products but ironically, studies have shown that this style is actually very poor in performance, especially in the recovery of fine gold. I wish that I had known about this stuff 40 years ago. Maybe I'd be rich now if I'd been using a better sluice box riffle design. Around 1901 many commercial sluice operators began using regular old angle iron segments as riffles primarily because they were stiffer and could span longer distances in very wide steel sluices without intermediate longitudinal reinforcing rods that were 20
Sluice Building sometimes necessary where inclined flat strap riffles were used. These operators and many engineers immediately noticed that this upside-down `L' shaped riffle created a very pronounced concentration effect on the downstream side, far in excess of that seen on other types of riffles. All types of arrangements were used over the years including experiments with various slopes as seen Figure 7-A,B, and C but perhaps by accident somebody happened to use what is called an `unequal length' angle iron where one flange is shorter than the other. This design worked so well that for all practical purposes this shape became the de-facto standard for riffle design for many decades in areas where fine gold recovery was an issue. This arrangement is illustrated in Figure 10. Notice that the horizontal portion of the `L' is shorter than the vertical leg. An optimized version of this style would have the riffles slightly tilted downstream at a 15-degree angle but reversing the tilt towards the upstream direction has been shown to be effective as well. (See Clarkson-Peer 10).
Over the later years observant engineers refined this design concept even further and the result was the invention of what we now call the `lazy-L' riffle that mistakenly is often referred to as being a Hungarian riffle as shown in Figure 11. The `Lazy-L' riffle combined all the best qualities of all previous riffle designs in that it functioned well in the collection of large heavy materials in the fashion of the old original riffles by simply blocking the passage of large nuggets in the pockets on the upstream face as well as creating a powerful downstream eddy vortex that concentrated heavy but smaller material in a low pressure zone on the downstream side.
Sluice Building The problem with this new and improved riffle however was that it had a tendency to pack-up rather rapidly so it needs cleaning far more often than more conventional designs or it ceases to concentrate properly and fines are washed out in the tailings, sometimes more often than when using less sophisticated riffles. As with most things about gold recovery almost everything seems to be a compromise depending on the particular characteristics of the source gravels and the size of the gold particles. (The best study I have seen to date on riffle design and performance is the one prepared by Randy Clarkson and Owen Peer for the Klondike Placer Miners Association in 1990. This thirty page study is available on-line from a variety of sources and I recommend that every serious prospector and sluice builder read and study it thoroughly).
Most of us small-time operators fortunately have it a lot easier than our larger commercial cousins and we can take the time to tend our sluices more often, classify our source materials down a little further and in general take more time in working our claims or favorite sites. For this reason the `Lazy-L' riffle design is almost perfect for a part-time operator who needs to recover nuggets as well as very fine gold. As good as the `Lazy-L' is it does have some potential problems and people who have never run a sluice with these riffles will have to learn a little bit more about how they actually work. They are far more sensitive to proper sluice setup and material classification than conventional riffle designs and this is the primary reason that so few sluice manufacturers have switched over to using them in their product lines aimed at part-time amateur prospectors. First of all since they are more efficient in general the box has to be tended and cleaned far more often than most of us typically do when we're running a box with more conventional riffles. A traditional slanted flat bar riffle system can be run for hours on 22
Sluice Building end and still function relatively well but a box equipped with `Lazy-L' riffles can begin to pack-up after only one or two buckets of material have been processed. This is the major reason that most dredges still use the old slanted-bar riffle systems. Figure 12 illustrates in detail exactly how a `Lazy-L' or so-called Hungarian riffle actually functions. Crude angle-iron riffles perform in the same manner and are almost as effective but sloping the vertical portion of the riffle seems to improve performance slightly.
As mentioned earlier almost all riffles, even the old original wood batten type riffles generate a negative pressure area on the downstream side but the slanted angle iron and later `lazy-L' riffles really take advantage of this phenomena and both are extremely effective at concentrating heavy particles on the upstream and downstream sides of the riffles. In operation this style of riffle is slightly more stable than the old slanted flat-bar riffle with respect to water flow and velocity in order to `work' properly. By `work' we mean to cycle material ahead of and behind the riffle through the action of `concentration' where lighter materials are displaced and replaced by heavier materials. The one area that needs to be studied in far more detail however is the shape of the rather short horizontal portion of the riffle that to a large extent determines the length and trajectory of materials that pass the riffle and enter the next low pressure separation zone created by the next riffle in the line.
Sluice Building Many people who build their own gear have found that altering the angle; size or even the shape of this small lip can change the characteristics of the low-pressure vortex significantly. This is something that we will be investigating during the testing phase of this project that will be published as Part III of the series. Figure 13 depicts a typical `Lazy-L' riffle with an up-sloped end as many people have been having good success with.
Figure 14 illustrates another adaptation on this same theme with a `Lazy-L' riffle that has but one possible shape of a compound angled end section.
Builders have also reported good results using this type of alternative riffle design so there is still much to be learned about various shapes. Another factor to explore is the spacing of the riffles and of special interest to me is the specific characteristics of both the thickness and shape of the `parting' edge at the end of a riffle of any type since I suspect there is significant information to be gained by looking at this single element with respect to `parting angles' as they relate to the separation of particles into the boundary layer between the laminar and turbulent flow. In the mid sixties I had an opportunity to work with a guy who had an almost dry claim in Nevada so he had to conserve and recycle most of his water. The sluice he built used segments of corrugated roofing and was extremely efficient as far as I could tell just from observation since it used very little water volume. As a result of that experience I'd like to take a look at several different types of `compound' riffles similar to that illustrated in Figure 15.
Preliminary and Interim Summation
So far I've learned that the successful recovery of fine gold depends far more on the accurate control of water volume and velocity in proportion to very discrete preclassification of source materials than it does on specific designs of riffle systems. In a way this was a disappointing finding as I had hoped initially that a `better' riffle design would be the solution to building a sluice that could recover both larger pieces of gold and the very fine stuff that many of us tend to accept as being a calculated loss that we have to live with unless we design our boxes just to capture small particles and forgo the capture of larger material. I will continue however to look for a better riffle design as I'm still convinced that this particular element of a sluice box is one of the things that we have the most control over. With respect to power sluices and dredges the problem seems not so much in trying to build more `stages' into the sluice but in attempting to develop ways to control the water flow in multi-staged boxes. A three-stage box with each stage receiving the same amount of water flow and velocity is actually nothing more than two extra boxes that aren't performing anywhere near optimum levels. It's a good idea but one that still needs a lot of design work.
Since I don't have access to a laboratory or special equipment the testing of riffles performed for this article are largely nonscientific and empirical but I believe they are as representative with respect to results as many more technical reports cited in the references.
Test Site The placer site selected for the testing is a large one-mile long gravel bar located on the eastern bank of the Bear River just outside of Colfax California. This particular spot was selected because amateur prospectors have used it for decades. I have used the site myself for the past five years. The site yields primarily very fine gold in flake form but also some nice small `picker' type nuggets. The typical disposition of gold in any specific spot on the bar is 1 gram per 15 gallons (dry) of raw freshly dug gravel. The gravel matrix consists of fine sands, light clay and small predominantly quartz and slate granules that range in size from 3/4-inch to 200-mesh packed into the large cobble stone (6" cobbles) bar. On a recent outing with over 60 prospectors spread out over the one-mile area all teams, mostly two and three person groups, recovered almost identical amounts of gold in the proportions mentioned above. For this reason I consider this site to be excellent with 26
Sluice Building respect to having a relatively consistent distribution of fine gold intermixed in the gravels. Based upon my 40-years of experience I have never found a site better suited for the conduction of recovery testing in a naturally occurring river bar.
Test Procedures The riffle tests will consist of two different levels. One will be gold recovery in the field at the Bear River site. The second level will be a controlled situation where concentrates from that site will be processed in a variety of sluices at my home where I can control the volume and velocity of water in each sluice and the classification of materials through various sieves. The test sluice box will have one side made from Plexiglas sheet so the concentration activity of the water flow and riffles shape may be documented and photographed. This is an ongoing project that will be conducted over the summer months for this season and I will be updating the material periodically if I find new information that seems important to us part-time prospectors. If you want to contact me my email address is [email protected]
Happy Prospecting, Gary
1. Poling, G.W. and Hamilton, J.f., `Fine Gold Recovery of Selected Sluicebox Congifgurations', University of British Columbia (1986). 2. Wang, Wenquian and Poling, G.W., `Methods for Recovering Fine Placer Gold', (1983). 3. Clarkson, R.R., `Gold Losses at Klondike Placer Mines', Klondike Placer Miners Association (1989). 4. Peterson, L.A., et al, `Evaluation of the Effects of Total Suspended Solids Levels on Gold Recovery in a Pilot Scale Sluice', Kohlmann-Ruggiero Engineers (1984). 5. Peterson, L.A., et al, `Evaluation of the Effects of Total Suspended Solid on Riffle Packing and Fine Gold Recovery in a Pilot Scale Sluice', Centec Applied Technologies (1986). 6. McCarter, W.A., `Placer Recovery', Klondike Placer Mining Association, Canadian Inspections and Exploration and Geological Services Division Presentation paper, (1982) 7. Hague, J.M., `The Recovery of Lode Gold in Jigs', Engineering. and Mining Journal., Vol. 141, No. 4, (1940). 8. Longridge, Cecil Clement, `Gold Dredging', The Mining Journal, (1905). 9. Poling, G.W, `How Riffles Work, Results of Yukon Study', University of British Columbia, (1982) 10. Clarkson, R.R. and Peer, Owen, `An Analysis of Sluice Box Riffle Performance', New Era Engineering Corp., Klondike Placer Miners Association, (1990). 11. Williams, David, `Georgia Gold Rush; Twenty-Niners, Cherokees and Gold Fever', University of South Carolina Press, (2003). 12. Phillips J. Arthur, Mining Engineer, `Mining and Metallurgy of Gold and Silver', F. and F.N. Spon, London, (1867). 13. Zamyatin, O.V. et al., `The Concentration of Auriferous Sands and Conglomerates', Moscow, Nedra press (1975). 14. Young, Otis E., `Western Mining: An Informal Account of Precious Metals Prospecting, Placering, Lode Mining and Milling on the American Frontier from Spanish Times' University of Oklahoma Press, (1982).
Sluice Building 15. Egleston, Thomas, `The Metalurgy of Silver, Gold and Mercury in the United States', (1887). 16. Madonna, James A., `Precious Metal Recovery Using Sluice Boxes in Alaska and Canada', School of Mineral Engineering, University of Alaska, Fairbanks, (1988). 17. Tourtelot, Harry A., `Hydraulic Equivalence of Grains of Quartz and Heavier Minerals, and Implications for the Study of Placers', U.S. Geological Survey Professional paper, (1968).