starting a build for a robotic dredger

[Fermion]

Hi guys,

I am an electrical engineer with a hobby of mechanical engineering. My wife is a software engineer who does most of the microcontroller programming and UI in our past projects.

We have been interested in gold prospecting just for fun for several years, and watching Bering Sea Gold has us thinking of merging some of our past robotic projects with the idea of gold dredging.

It is somewhat hard to calculate if this is a viable project. The fun factor would convince me to build it if I have a chance of at least recouping materials costs. I already have a very large cnc bedmill and lathe in my shop, plus plenty of electronic test equipment.

The robot would allow the dredge hose (considering a 6 inch hose) to be remotely operated underwater with several cameras and lighting to give feedback to the user on the surface. Ideally it would be very suited for dredging under the ice as the entrance hole would not need to be much larger than 12 inches and possibly could be made with an auger. This would save time when spot checking several sites. All of the exposed parts of the robotic section would be made of stainless, delrin and rubber.

I intend for the robot to be controlled by a human but will have some canned sweep programs to allow it to run semi-autonomously. It should be able to grasp or nudge boulders up to about 300 pounds.

One of the sticking points is actuators. I am unsure right now what would be the best choice. I have considered hydraulic, pneumatic, hydraulic using pneumatic cylinders, and electro-mechanical (ballscrew/servomotor) actuators. It might be best to proceed on this project by developing the actuator and cycling it under load at depth pressure in salt water for several hundred hours to make sure it is robust. Once the actuator is proven, the rest of the robot is somewhat trivial as is the programming.

Anyway, if you have any ideas or suggestions as to approaching this project, I would be happy to hear them.

 

Amazon Forum Fav 👍

Treasure of Victoria Peak - Grab it through Amazon!

[Fullpan]

Very interesting concept - "sounds good on paper". I like the idea of starting small, like they do on "Myth Busters" show to demonstrate "proof of concept"
and "repeatability" before gearing up. Remember that at least half of the effort "down there" consists of "rock chucking", that grasping odd shaped cobbles
weighing from 2 lbs up to 60 lbs and more, and chucking them as far away from suction nozzle as possible. Hope this helps.

 

[Fermion]

The plan for the nozzle end is to have cross bars to prevent large rocks from getting stuck inside the hose. I have considered having a second set of controllable pincers that could open up and grasp cobbles then toss them....or it might be good enough to just bat them aside with the nozzle.

Can a diver move a 200 pound cobble? The robot might have some advantage there, in addition to never getting cold, tired, or running out of air.

On paper it sure sounds fun...maybe if I built it I could get a TV show....RobotFathers or something...

 

[improvmechanic]

So you're talking about building a submersible vehicle then? Presumably launched from a boat on the surface? Sounds interesting.

I don't know if it would be practical to move 300 lb stones, but certainly the ability to push 60 lb boulders out of the way would be a plus.

Just to satisfy my own (now extreme) curiosity, would you be attempting to build essentially a submersible type vehicle or some sort of crawler?

 

[Fermion]

So you're talking about building a submersible vehicle then? Presumably launched from a boat on the surface? Sounds interesting.

I don't know if it would be practical to move 300 lb stones, but certainly the ability to push 60 lb boulders out of the way would be a plus.

Just to satisfy my own (now extreme) curiosity, would you be attempting to build essentially a submersible type vehicle or some sort of crawler?

Well, I was going to keep the actual design secret, but what the hell, everyone always claims they have some super great idea but never reveal it so here goes:

I plan on building a snake with at least eight articulated sections controlled by the actuators. Probably three actuators per section spaced 120 degrees apart. The six inch dredge hose would run down the middle of the snake and there would be an outer protective shield (maybe a 10 inch hose). The articulated section would be about 20 feet long and would be capable of pulling itself along with the rest of the 6 inch dredge hose leading to the sluice above. There would be no exposed wheels, legs, or anything else to get caught on stuff...just the relatively smooth body of the snake. Near the head of the snake which would be the dredge nozzle, I would have at least 2 cameras mounted on opposite sides of the snake along with a ring light of high power leds.

All in all, it would look something like the spice worms in Dune and probably be scary as hell on land, rearing up 20 feet in the air...especially if I add the pincers at the nozzle to grasp rocks.

 

[improvmechanic]

Eight points of articulation?! Incredibly ambitious! If you start work, I don't think you have to worry about people here stealing your intellectual property. And I'm no engineer, but I am a bit of a mechanic. It sounds like a heck of a stretch.

I'd watch out for hydraulic and compressed air actuators. This thing sounds a little complex for those. You could wind up with a hydraulic nightmare. I like the idea of electronic actuators, but I'm not sure you'll be able to get the strength you're after. It does sound like it would be simpler to maintain, and you don't have to worry about hydraulic fluid or air lines.

Have you seen the Keene Nessie model 8DX? That's about the closest parallel I can draw. You may get an idea or two to adapt from that design.

All in all, sounds like you've got one hellova plan there bud. I'd love to see this in action. If you want a sounding board for ideas I'll offer my services.

 

[Fermion]

Eight segments in the snake seem like a reasonable amount to allow it to work a significant area before it needs to drag itself and the remaining hose around. When it needs to move a greater distance than the 20 foot reach it will open the pincers at the nozzle, plunge it's head into the sand/rocks, and arch the segments. This should have the effect of pulling the trailing hose along the path of the snake. A slithering motion might be possible too, but could be more problematic depending upon the topography of the bottom.

I am also leaning toward the electro mechanical actuators mostly because I have a lot of experience with brushless servomotors, controls and ballscrews. I think I can make a relatively small actuator sealed inside a flexible rubber boot that could provide around 300 to 500 pounds of force. If the actuator can push as well as pull, then with three actuators per section I think the strength of the snake joints would be enough to do what I want.

I may make a scale model of this with airplane servos just to test the movements and forces. I certainly appreciate a sounding board for my ideas. I also have very little experience ocean dredging and would love to hear feedback from divers too.

 

[MadMarshall]

Just plain crazy!! How I admire the visions of men and their ability to conquer barriers I think I'll stick to my hammer and pan

 

[Lanny in AB]

As I've dredged quite a bit, you'll need some way to prevent nozzle clogs. Bars across the nozzle won't do it, particularly in certain conditions it will be much worse than in others, as suction will slap flat rocks, and other rocks will pile on those bars and then accumulate more until the nozzle can't suck anymore. So, if you can come up with a viable way to remove the rocks that are blocking the bars on a regular basis--you'll be in business.

I remember once a guy came to me with the idea that he'd just design a claw (parallel bar design in a claw shape) that would stop the nozzle from clogging. We were working in slate, and as you know, slate comes off in lots of flat pieces. That's not a viable design for slate for sure--it just clogs quickly on the end of the nozzle preventing anything from going up the nozzle.

When I was in Alaska, I was talking with Perry Massie about a similar idea: we were working on Mystery Creek (the sluice was loaded with gold top to bottom on that day's clean-up--it was a gorgeous thing to see), and I was quite inexperienced. They were digging through a gravel seam, and the heavy gold was in the broken bedrock underneath. I suggested that maybe they could hook up some kind of dredge nozzle that they could attach to the bucket underwater and then somehow after the bucket got through the gravel, they could use the dredge nozzle to suck the gold from the bedrock as they drug it along. I found out what a kind person he was that day as he considered my idea carefully and then explained how the nozzle would get jammed as there would be no way to classify the rocks entering the nozzle. I mean, he's a mining engineer, and I was a green, but eager; however, I found out later that I was totally out to lunch (as I found out when I started to dredge--it all came together in a hurry for me then).

Good luck, and if Keene has found a way to do it robotically or remotely, there must be a way.

All the best,

Lanny

P.S. I see that you're thinking of pincers at the nozzle, which would avoid ramming rocks in the nozzle, so does that mean that the suction will be off until you're ready to dredge? If so, how will you handle rock jams within the nozzle or the hose when they, inevitably, occur?

 

[bill-costa rica]

I have thought about building a robotic arm dredge.. a 6 inch is way 2 small. you would need a 16 inch hose. the gravel intake necked down to 14 inches. that would stop most plugs. for plugs at the gravel intake you would need a hydraulic ram to push the rocks off the intake. another ram would be needed to push rocks back out of the jet. this is where most plug up occur have something like a back hoe arm to move your hose would be very sweet. you would be limited to about 20 feet from the surface of the water. getting a permit to run it might be a stumbling block and maybe not. fun to dream anyway. all the technology is there, all you need to do is customize it a bit

bill –Costa Rica

 

[Fermion]

Ok, thanks Lanny and Bill for the heads up on the rocks/clogging problem.

How do divers clear the nozzle with a 6 inch dredge? If the nozzle had a pair of crossed fixed bars in an arch across it, creating four openings that would allow only ~3 inch rocks to enter the hose, would larger flat rocks still stick to the openings? Do you think just scratching the nozzle across the ocean floor with the robot occasionally would be enough to clear stuff stuck to the bars or would something like the moveable pincer jaws in addition to the bars help to clear the nozzle?

I was thinking 6 inch because my budget is not unlimited, and trying to build a 24 actuator snake on a 16 inch dredge would be well beyond what I can imagine spending.

As far as clogs happening past the bars, how often would that occur? What do divers currently do when that happens?

 

[improvmechanic]

I like the idea of the bars. Keene has a suction nozzle with a "suction breaker." It's a rubber flap over a screen further up the nozzle. When the flap opens the water rushes through the screen and the rocks fall off the end of the nozzle. Should be simple to remotely actuate

 

[Lanny in AB]

Ok, thanks Lanny and Bill for the heads up on the rocks/clogging problem.

How do divers clear the nozzle with a 6 inch dredge? If the nozzle had a pair of crossed fixed bars in an arch across it, creating four openings that would allow only ~3 inch rocks to enter the hose, would larger flat rocks still stick to the openings? Do you think just scratching the nozzle across the ocean floor with the robot occasionally would be enough to clear stuff stuck to the bars or would something like the moveable pincer jaws in addition to the bars help to clear the nozzle?

I was thinking 6 inch because my budget is not unlimited, and trying to build a 24 actuator snake on a 16 inch dredge would be well beyond what I can imagine spending.

As far as clogs happening past the bars, how often would that occur? What do divers currently do when that happens?

The bars (curved or arched or not) will stop rocks--including flat rocks (anything round over three inches as well). The rocks will then stop other rocks--then you have no suction. (While a person is dredging, you're constantly keeping the nozzle clear--the nozzle is constantly grabbing rocks bigger than the intake as well.) The six-inch dredge nozzles can have a flap on the side to open with a pull handle that breaks the suction so that you can then remove the rocks--or your arm. The amount of suction created by a dredge is incredible. It will slam a rock across the nozzle with a tremendous smack! It will pinch/damage your fingers or hand.

If you're opening a flap remotely on the side of the nozzle, then other rocks will rush in to the side and jam it, unless you have cameras mounted so that you can raise the nozzle clear of all the cobbles each and every time, or you raise the nozzle clear every time you have a jam. That's probably why the person is a key in the operation.

Remotely trying to keep the nozzle clear would be a challenge, but it may be possible with some of Bill's suggestions. It does sound like a very expensive build though.

To remove jams, you swim the line--trying to see where the jam is in the hose, then you use a rubber hammer to carefully smack the side of the hose to get the jam to move, while the dredge is still running under suction. If the edges of the rocks are sharp--you puncture your hose--then you have another problem to deal with--loss of suction, and if the hose rolls above the surface--you'll suck in air, which creates another problem. If the clog is on the front/face of the nozzle, you try to pry them off, or you use the suction-release flap on bigger dredges; or, you signal the surface tender and have him kill the dredge, and then the rock/rocks will drop off. If the nozzle gets a jam in the bend, you can try killing the dredge and hope it will drop free, but often it requires a bar and a hammer to work it loose. If rocks jam in the jet (just in front of the flare/slick-plate), the dredge tender (dredger's mate) runs a rod into the flare, down to the jet and tries to work the jam out/bash it out. If that doesn't work, you kill the dredge, take the suction line off, and then you might have to get your arm in the jet to try to work the jam out, or you get a bar and try to work it loose/drive it out from the open end where the suction line was attached.

You have to remember that those rocks go in under a great amount of suction and force, and when they jam tight--they're really jammed!

All the best,

Lanny

 

[Fermion]

I obviously may have underestimated the work a diver does to keep the nozzle clear. It sounds like half the time is spent unclogging the nozzle or removing jams.

I guess a person can't know everything...probably not a lot of divers know how to build a brushless motor controller :-)

This is a good reason to discuss ideas on a forum, it brings these issues to light.

I will have to think on the rock jamming issue and see if there is anything clever I can come up with other than constantly cutting suction and having the snake rear up and shake off the rocks or scrape them off against a much larger rock. I had assumed the most exhausting and hardest part of dredging was the constant dragging of the nozzle from area to area and sweeping it for hours. If you have any more suggestions or workarounds for the nozzle I would be glad to hear them. Thanks.

 

[heartslayer2001]

keeping the nozzle clean is a simple task in this, the actuators would need to be strong enough to move against the currents of 20-25 ft of water. As far as boulders if you make the nozzle right you need not worry about them. think about a cone shaped nozzle taking only dirt and small rocks. consider cone 18'' tall, stainless 1/4'' attached via rings, holes run 1'' all up and down you can get into small spots or dig deep with it. just a thought, also 8 joints , try one 10 ft then 3 at 2.5ft.

 

[oldbrundogg]

I don't really have any right to be posting on this thread except that I think a lot . What about if the nozzle had like drain cover with something a kin to a windshield wiper that spins around the nozzle cover to knock the stuck rocks off periodically actuated from the controller above. Never stopping the suction.

 

[Lanny in AB]

keeping the nozzle clean is a simple task in this, the actuators would need to be strong enough to move against the currents of 20-25 ft of water. As far as boulders if you make the nozzle right you need not worry about them. think about a cone shaped nozzle taking only dirt and small rocks. consider cone 18'' tall, stainless 1/4'' attached via rings, holes run 1'' all up and down you can get into small spots or dig deep with it. just a thought, also 8 joints , try one 10 ft then 3 at 2.5ft.

The nozzle grabs the rocks, and it can't sort them--I need more feedback from you on what you mean that keeping the nozzle clean is a simple task. It's not boulders that are the common worry--they are a worry--it's the assorted cobbles that cause the majority of the problems, and you have to move them to get at the gold beneath them, between them, and then the rocks often have to be pried apart they are so tightly jammed--especially when you get close to bedrock. We're not talking about loose, easily moving gravel. Stream packed material is hard to work. If you have holes drilled in a nozzle, and you've got the suction of a six-inch behind it--the rocks will stick like ticks to those holes.

All the best as you try to work this out,

Lanny

 

[soupfreak]

maybe instead of a cone with holes, a cone in the form of a grizzly like on the Bazooka Gold trap sluice. Then bars on the inside of the nozzle that could "scissor" out to clear obstructions, and maybe even be used to dig in to crevices.

As for keeping the hose clear, maybe like a 1" flat hose inside the dredge hose that could be deflated and re-inflated to work any clogs loose.

I'm sure these ideas are not without their drawbacks, but maybe a start in the right direction.

 

[Lanny in AB]

I have an idea that would get lots of dredgers in on your conversation--transfer your questions and ideas to the dredging forum and give them a shot at this. I'm sure with more "dredging" operators in on the discussion, it should clear some things up for you. I've dredged for many years, but I only represent one point of view.

All the best,

Lanny

 

[AzViper]

Ok, thanks Lanny and Bill for the heads up on the rocks/clogging problem.

How do divers clear the nozzle with a 6 inch dredge? If the nozzle had a pair of crossed fixed bars in an arch across it, creating four openings that would allow only ~3 inch rocks to enter the hose, would larger flat rocks still stick to the openings? Do you think just scratching the nozzle across the ocean floor with the robot occasionally would be enough to clear stuff stuck to the bars or would something like the moveable pincer jaws in addition to the bars help to clear the nozzle?

I was thinking 6 inch because my budget is not unlimited, and trying to build a 24 actuator snake on a 16 inch dredge would be well beyond what I can imagine spending.

As far as clogs happening past the bars, how often would that occur? What do divers currently do when that happens?

Could you not put in place a low suction pressure switch so that a loss of suction due to build up at the nozzle would shut the pump down for a few seconds so the nozzle would clear it's self? Your correct in that you would need to limit the size of rocks being pulled into the hose so they do not get stuck and with bars across the nozzle will clog easier.