I will be building an exhibition machine that will require an excavator type arm (big arm, boom, bucket- about 26 ft total length) carrying an 1800 lb load where the bucket normally would be. However this load will need to be accelerated beyond the normal operating parameters of a standard arm. The load will need to be launched into the air about 30 ft. From beginning of lift to release about 1-2 seconds, ~3g acceleration. It will be caught, some energy regenerated as fluid is pumped back to the accumulators, and the cycle repeated. I have located high speed cylinders and fast servo valves but have not seen booms or arms designed for these stresses. Before I attempt to re-engineer the arms I was hoping to find examples of booms or arms capable of this extra duty. Any ideas? Thanks all
Pretty hard to envision exactly what you're doing and the various functions. 1800 pounds ain't much and I would think as long as you have a sizeable unit and ramp the flow up and down with valving, cushion accumulators on the cylinders, internal cylinder cushions, etc. you shouldn't have much trouble. Of course if anything does fail it will be spectacular.
Thanks Ian- I guess by choosing a larger arm than would normally be required by that load, the increased stress on the welds by the higher acceleration would be compensated somewhat.
The digging forces seen on a daily basis would far exceed what you plan on doing I think. The rapid acceleration/decelerations on a machine with a hammer would likely be comparable albeit a shock load vs gradual loading.
It would be nice to have a little more info, what size and tonnage machine. A boom/stick set up for a front shovel, a feller buncher or logger type boom, or a scrap handler/demo type boom. 1-2 seconds would be fast launching 1800lbs
This will be a crowdfunded project and the proposed machine will be made public in a few weeks. But I can say that will be 70ft tall, on a static base of two legs, will have two arm sets each with a "hand" for grasping and releasing. It will literally be juggling three large objects. There will be an operator in the "head" with haptic feedback control of the action under a protective rollcage and thick polycarb. dome. The arms will be either custom built or will use off the shelf parts that are adapted to the task. I once built a large pipe mover with 4 hydraulic wheel motors and 4-wheel steering but this is new for me and that is why I came here looking for suggestions. Thanks
If you haven't already, I would think about building maybe a quarter scale one with remote controls, before you build a potentially lethal catapult? Oh, and I don't care what the hell you build the operator's cage out of, if you hit it with an 1800 pound object, it's likely he'll be dead, or wish he was. I'm not trying to be a jerk here, I'm just saying that this scares the he** out of me!
Ian, You are right about the potential of this thing and I will build a proof of principal 6ft tall one arm unit to prove out the controls. The big one will take lots of funding and much engineering, including safety. Among my heroes are Isambard Kingdom Brunel (Great Eastern) Roebling (Brooklyn Bridge) Richard Branson (private space vehicles) All chance takers. Of course I will be the first to climb into this thing to test it.
Is there a general formula for required HP? I can estimate GPM and PSI for steady state running including charging the accumulators. I would like to have 20-30% over capacity.
Pump horsepower = q p / 1714 where q = required pump capacity (gpm) p = required pressure (psi) This is a basic formula assuming 100% efficiency which never happens. Estimate 80%.
Nothing personal, but I think you're in way over your head. If this is for real, you need to start by finding a servo-hydraulics expert and see what it will cost to do what you want to do. I expect that would be much more expensive than you think. Your requirements are way beyond the realm of construction equipment.
[ Yes, that's a fair description of my ambitions in this case (actually 64ft over my head). But I like the challenge. Anyway the budget will be about $1.2M to cover those consultants and certified welders, etc. The reason I want to borrow from construction equipment design is that it has been hard tested and improved by failures and disasters and fools like me who stretch the limits.
If you trying to accelerate the load of 1,800 pounds 30 foot up in two seconds it will take roughly 50 hp. One second will take 100 hp. This is ignoring weight of the boom and stick, losses for mechanical friction, the fact its actually moving in an arc, and etc. A standard boom and stick will off a 12~20 metric ton machine should be able to accomplish what you need. I have a Deere 490E boom and stick that should fit the bill for the linkages you need. Assume 2,500 pounds for boom and 1,500 pounds for the stick. Moving 1,800 pound load quickly will not overstress the linkages. Digging and rock hammer is much worse. However, you will need much more than an applied 50 hp to accomplish your goals if your doing this with a standard excavator hydraulic system. Moving something that far that fast with linear hydraulics will have very high fluid friction rates and will need very large pumps and valves. If you show a sketch of what you want to do, it should help get you further along.
Thanks IHM, It's good to get some basic numbers and estimates. The animation for this is almost complete, then I will post the links on this forum. Meantime for the proof of principle 1-arm unit I am seeking at least one used servo valve with amplifier/controller so I can demonstrate the control system. (starving inventor $ budget til funding kicks in)
I think that it would be much more expensive than you are thinking. To juggle objects as a person does, the objects have six degrees of freedom (of movement). The juggler's hands have six degrees of freedom in manipulation of the objects. This is nothing at all like construction equipment. Scaling factors will hurt you. Strength increases as the square of size, mass as the cube of size. A one tenth size proof-of-concept model won't mean anything, one half or one quarter size might (there is still an eight-fold increase in mass going from one half size to full size)(and and eight-fold increase in price; price scales as mass). Tossing the objects up 30' high, they accelerate over ten times more distance than tossing them up 3'. The response frequency of your control system needs to be very high, but the natural frequency of the structure needs to be higher, and that will be very difficult with the structure size (large mass and long levers) that you are talking about. To put it simply, to juggle objects the ends of the arms need to move as fast as your hands do. When you want to juggle something 30' up, they need to move even faster. When you apply force to the arms they will both bend and move. If they bend and bounce very much, you have no real control over the motion. Ask someone like Peter Nachtwey http://hydraulicspneumatics.com/author/peter-nachtwey how much the servo-hydraulics will cost. I have seen systems with the power and speeds you are talking about; they are in the aerospace realm. (The power and speed requirements you are talking about are similar to that of the Space Shuttle main engine steering; the speed of control surface actuation but with ten or a hundred times more power).
Jim, Here are some informal ballpark calcs done by a friend using SMath. http://proto.dangyro.com/objcalcs.jpg If you see orders of magnitude errors, let me know. I am assuming some regenerative energy recovery on weight catching. Thanks for your input- Between your high caution and others' "no problem", there is a doable machine. And it will have all necessary degrees of freedom. The stresses are the main issue. And since it will be crowd funded, the budget may expand. The Monster Truck people alone might want to fund it. Dan
Dan, If this is for real, and I don't know if it is... you need to ask a real engineer to assess the idea before you waste any real time or real money on this idea. The scratch sheet calcs you have mean nothing. Sorry. The structural, and servo-hydraulic control, impedance problems are huge. This following is a thought exercise to elucidate the problem: I imagine that you can hold a tin cup in each of your hands and juggle three balls, by catching and throwing them with the cups in each of your hands; I've seen that done. Now imagine putting the two tin cups on the ends of two fishing poles, and holding the fishing poles from the hand grips. I doubt very much that you could juggle anything from the two cups at the ends of the fishing poles, by holding and moving the poles from the hand grips. The response frequency of your neuro-musculature is higher than the natural frequency of the fishing poles; your hands move quickly but the cups at the ends of the rods wobble around, you get the idea. The natural frequency of 26' arms throwing one ton weights will be lower than fishing poles with tin cups attached to the ends of them. I don't think you could do it for $14 million. I've seen structural and dynamic test rigs that had close to the servo-hydraulic power/speed density you are talking about; $1 million would buy some of the parts of such a system, and would only buy a very small part of the engineering costs! Building a very complex, large and powerful machine from scratch is *very* expensive. Your in the range of the prime mover power and the flow rates of a hydraulic mining shovel. The following is a paraphrase of what a very good rocket scientist wrote to explain the relative cost of different technologies: You want to use an excavator stick and huge cylinders and accumulators to throw a one ton mass up into the air... Doable. You want to use the arm and hydraulics and some servo dynamic control to throw the one ton mass straight up into the air and catch it on the way down... Bring a *big* check book. You want to have two arms on each side of a tall structure and want to throw the one ton mass from one arm and catch it in the other arm... Do you *like* pain? You want to *juggle* three one ton masses from the two arms, with hapsatic control of the arms... Ah..! You *DO* like pain! PS: If this is a burning man thing, ask if they'll let you do such a stunt. The BM people pay a lot of money for a very large dollar amount bond that will pay for the remediation of any environmental mess that is left behind. Without expensive engineering and failure analysis, you can't provide any meaningful assurance that you won't spray hundreds of gallons of hydraulic fluid onto the playa. Dry-break separation of the hydraulics with structural failure is probably impossible for your machine.
PPS: To try to put this into an easy understanding. You are talking about juggling. It takes a lot of practice to learn. That means it takes a lot of neuro-muscular feedback to learn. That means it takes very precise neuro-muscular perception and precision of motion of the hands to learn to juggle. Learning to juggle one ton things thrown 30' up by 26' arms will take greater precision of motion than the precision of motion needed to juggle with your hands. And that precision of motion is by way of servo-hydraulic control of very large masses and moments. And the juggling 'hands' will be some distance peripheral to the juggler. The fidelity of motion of the 'hands' will have to be very, *very*, good to allow you to learn to juggle with your machine. That precision of motion is *way* beyond construction equipment, and, in displacements. it is way beyond the Space Shuttle main engine steering system.
Thanks for the reference to Peter Nachtwey - excellent real world hydraulics pointers you don't learn in school. Reminds me of a classic on steel fabrication by Carroll Smith: Engineer to Win. He built race cars and their frames, siting subtle tricks to avoid stress cracks and welding/machining errors. Your analogy of a fishing pole to a box welded boom is flexing it a bit. Fortunately there are active/adaptive oscillation damping solutions in use for decades for boom arms. One of many studies and applications: Active Vibration Damping of Mobile Hydraulic Machines http://www.ccefp.org/research/thrust-3-effectiveness/project-3b3 Re: the Shuttle engine- 2 axis actuators tilting a gimbal +/- 10.5 deg. against 500,000lb thrust relates to this how ? Juggling response time: remember scaled up, free fall lasts longer, there is more time to correct. And I may add cameras to the hands to help solve the parallax problem of positioning the distant hand to help catch the object. As to expense, many equipment mfg's donate parts and expertise to such projects especially in support of nonprofit orgs which ours is. We'll see. I sense you are not a typical "naysayer" but that you are trying to ward me off possible disaster. I intend to pursue it so I would welcome insights as to how it might be made successful. (I fully absolve you of all liability for information leading to injury, death, embarrassment etc.)
As an alternative to the traditional hinge/hydraulic-cylinder type joint, you may want to look at these rotary actuators; http://www.helac.com/ I have no affiliation with this company, other than I've had experience with their tilting quick hitches and found them to be strong and reliable.
I'll leave the physics to you, but the obvious issue I see with your calculations are, assuming .5 (50%, 1/2) regeneration, and 35' vertical acceleration distance. If you have a 26' arm the acceleration distance will be closer to the radius, right? or is this going to play catch, overhand? cause juggling underhand isn't impressive enough? Secondly, 50% regeneration is unrealistic, maybe in some other application but I'll bet it's not going to be economical (compared to adding more power) with hydraulics with such an intense cycle. Have you considered other regeneration technologies (pneumatic, flywheel/CAPACITOR) with hydraulics doing the "fine control"? An alternate power storage might allow a longer wind up, fast pitch softballish.
Delmer you are right- 50% recovery is too optimistic. The calcs were done by a young fellow studying for his civil engineer license and he does not have a background in hydraulics. I was mainly trying to get at an estimate for the required horsepower to run this. This will use the lowest stress juggling which is probably the "cascade" pattern- underhand. The falling object will decelerate and dampen the bounce as the arm pumps fluid back into the accumulators. I like your idea of alternate regeneration methods. I'll look into it.
Dan, The point of the fishing pole is to give you an idea of impedance. You can position the tip of the fishing pole very accurately, or you can move the tip of the pole very fast, but you can't do both at the same time. The same is true of an excavator; you can gently touch something with a bucket tooth, or you can swing the bucket fast, but you can't touch something quickly, (...tap...tap...tap...) as you could do with your fingertip. The shuttle engine steering has very high precision with very high forces. When a certain amount of movement is commanded, that amount happens in an known amount of time, without undershooting or overshooting or hunting or wandering around. At a very high frequency. And it takes into account the elasticity of the actuators and the structure. And it is very expensive. Your control system will need the same precision. When you move your hand and then stop your hand in another position, the hand on the 26' arm will have to move some exact distance and then stop, exactly in the new position that your hand has commanded. *Think about that* Some kind of feedback is required to maintain accurate positioning. This is a *very* big deal. You crash when your actuator position drifts away from your control position. Your hand could move *faster* than the arm can keep up with, like moving the handle of the fishing pole faster than the tip of the fishing pole can keep up with. Then what happens? Do you have artificial force feedback, or hard velocity limits, in the controls, to keep your hand movements within the control performance envelope (that is something that is done, but not easily... haptic feedback, written into computer code?) With construction equipment, your hand moves a control to move an actuator (one or two axis). *You* are talking about moving you hand, and having the juggling hand move proportionately the same movement (with some kinds of haptic feedback). Think about that; You need to map your hands' control space positions into the real space positions of the juggling arms' pivot rotations and cylinder extensions. And code for two integrations of the positions equations, for accelerations, for the 'force' feedbacks of the haptic controls. (etc... there is *so* much more, but I'm not being paid for this...) I can imagine that you could build a machine that could toss two objects into the air. (How do you toss the third one?; two in one of the hands andr throw only one.. or pick it up from somewhere...) Maybe a machine that could toss three things in to the air and crash into them with the launching arms. I can't see that you could build a machine that can *juggle* three one ton objects with 26' long arms that can toss them 30' up, catch and grasp them on the way down. and sling them back upwards and release them on a trajectory from which they could be caught again. Not for $1.4, or $14 million. But I could be wrong. Jim
I don't think that you can let the falling object crash into a stationary hand... and then catch the object. The hand needs to be in a synchronized motion, to catch the object. Regeneration only adds very much complexity, at great expense. Motive power is always cheaper. You really need to get a grasp of the motive power and fluid flows that your project will require.
For booms, look to logging equipment, Madill loader booms, Pierce Pacific Manufacturing booms and sticks and other road builder / logging stuff.
About special booms and hydraulics the best i know is LUYCKX the Hitachi dealer in Belgium , nothing seems impossible to them , they manufacture booms and equipments to the customers specs