PDA

View Full Version : Why doesnt no one use hydraulics?



El Espada
01-23-2008, 11:41 PM
Just curious why doesnt anyone use a hydraulic system to help move there bot? Done a lot of searching and havent found nothing on the matter. Well, found a lot of theory but no execution.

DresnerRobotics
01-24-2008, 01:04 AM
Size would be my first guess. Cost would be the second.

You'd have to build a pretty large bot to use/need hydraulics, considering the compressor and control setup for them. It's just not practical given the smaller sized servos and actuators available, unless of course you're trying to build a full sized mech :p

Zenta
01-24-2008, 10:14 AM
Hi!

I don't now about humanoid hydralic driven bots. But I'm pretty sure this walking tractor use hydralics:

http://www.youtube.com/watch?v=CD2V8GFqk_Y

Very cool!

-Zenta

jdolecki
01-24-2008, 05:59 PM
Allthough there is no such thing, I think i would call it "Micro Hydraulics"

You would have to design all your own Micro pumps, cylinders and valves.

The smallest Hydraulic pumps ive seem were used in Dock levelers.

Just a thought.

elcray
01-24-2008, 06:32 PM
I suppose on the mini or micro level, pneumatics would be a safer bet than hydraulics. Not as my power as hydraulics, but at the smaller scale, the power isn't as important. And, your power could be a small CO2 scuba tank. Just my 2 cents.

kdwyer
01-25-2008, 08:08 AM
A few reasons:
1. Hydraulics are expensive.
2. Hydraulics are (or can be) extremely powerful, ie. dangerous.
3. Hydraulics leak.
4. The motor(s) used to run the hydraulics could be better used directly as electrical actuators.
5. Hydraulics are expensive.

At 3000 PSI, hydraulics are positively hazardous. And it don't say "Ooops" or "I'm sorry".

For stationary bots, hydraulics IS an option. Materials handling etc. works great. Making the whole thing mobile, well...

vcambri
01-25-2008, 08:14 AM
And what about the friction? In large systems, the relative loss of power due to friction is relatively low. When you go to "miniature hydraulics" my guess is that friction in the pistons would play an increasingly important role. Just a guess though. I'd like to see more about this.

Matt
01-25-2008, 10:48 AM
Sort of a side comment. IFI actually has a decent sized (meaning small) Pneumatic system. The teams in the FIRST competitions use them for things such as raising and lowering gates, large claws, etc.

the images on the page seem to be down at the moment, but here is the link.
http://www.ifirobotics.com/edu-pneumatics.shtml

El Espada
01-25-2008, 10:54 AM
Ok to clarify!! On the mini and micro frontier servos are the way to go. But im talking about bigger stuff. Like say a crustcrawler that u can ride on! You guys do some really awesome stuff with the mini's and micro bots that I cant see why you cant transfer to a larger scale. I mean come on u have done all the programming, made a design, but why not go big. Aside from the cost of course i really see hydraulics being the only feasible option to make this work. For example http://www.neogentronyx.com/. I know 99% of you guys have seen this but this guy started backwards. No programming and no small scale design. For some of you the only challenge a project of this scale would be just enlarging your creations. Right now as far as i know this thing hasnt walked yet! But you guys could do something that can awe people. I know the post is kinda deviating from the topic but how would one use hydraulics for projects like this? I mean I just got into the whole scene of trying to build a robot this last year but i have researched this for a long time and i just curious why the guys that actually made some progress hasnt tried going big yet?

jdolecki
01-25-2008, 11:58 AM
My comments about Micro Hydraulics we said to provoke thought.

I understand that large scale Hydraulics are expensive and costly, But has there been any research into micro Hydraulics?

If we have mini cylinders that have .5 to 1.0 inch stroke how big of a of a pump and how much pressure do you really need.

I dont know if they would be more efficent or cheaper than servos but if no body has ever done it before then how can we really decide of its better or worse.

Wouldnt it be nice if you humanoid robot had a Hydraulic hand and in combat it could just crush the competors frame or electronics or mabe even clip his wires?

What about a human Prosthesis that uses the Heart (how much pressure is ther in the blood stream) and blood to powerer damaged/replacement muscles.

Is it possible?

Who will be the first to explore this technology.

Alex
01-25-2008, 12:12 PM
why not go big

You already answered your question - cost:D Cost is the biggest factor on why we don't have these systems available everywhere you go. The technology is obviously there to do this, but who is going to pay $400k+ for a crawler (one that doesn't look like it'll kill you if you get in it) that they can ride? More importantly, what venture capitalist is going to fund a project like this? What is the feasibility of having one of these other than just fun?

Just my $0.02.

Oh yeah, Anybots does some really cool work with robots and hydraulics too:

http://www.anybots.com/

We finally got to meet them out at RoboDevelopment. Really cool guys!

kdwyer
01-25-2008, 01:43 PM
The big problem with hydraulics is power. A mobile bot that primarily uses hyd. will need an internal combustion engine, generator, and hyd. motor. Notice, 3 components. Lotsa parts.
Electric motors are small enough and strong enough that hyd. sorta takes second place. Weigh 3 or 4 wheelchair motors, and compare to a few hyd. cylinders. And reservoir. And pipe/hose. And fluid. And connectors.
Electricity is just a cleaner quicker way of delivering power. Hyd. is usually much more powerful but doesn't scale down well. Electricity also provides feedback, hyd. just leaks 'n fails.

Matt
01-25-2008, 03:22 PM
I think another issue is the behavior of force in hydraulics. It's very intense isn't it? When you add pressure that lever/joint/appendage will suddenly have great force being applied to move and then zero once it gets there. It's a pretty violent way of moving around I'm thinking. Maybe I'm wrong, I admit to knowing almost nothing about hydraulics, so I guess I'm asking.

Here is Dexter from anybots, it's hilarious watching one robot poke another trying to knock him over.


http://www.youtube.com/watch?v=_CQ5AKaEi3U

Dave
01-25-2008, 04:15 PM
I love this thing:


http://www.youtube.com/watch?v=b2bExqhhWRI

BigDog is a hydraulically actuated quadrupedal robotic pack animal. It has some impressive dynamic walking skills. My favorite part of this video is when the guy kicks it and it recovers without falling over. I also like when it's bouncing around at the very end of the video. It looks so happy :). Like kdwyer pointed out, an internal combustion engine is necessary to make a system like this mobile. As a result, BigDog isn't exactly stealth.

Hydraulics and pneumatics are really cool, but not terribly practical for mobile applications. You've got the hydraulic BigDog, which consumes gas and sounds like a swarm of angry weed whackers. Pneumatics? I think every pneumatic bot I've seen has had a trail of hoses coming out of it. Never seen one that actually carried its own air supply. I've also been told that the compressibility of air makes writing pneumatic control code a pain in the ass.

JonHylands
01-25-2008, 04:16 PM
Just a note...

Dexter is not hydraulic - those actuators are all pneumatic...

- Jon

Alex
01-28-2008, 11:49 AM
Dexter is not hydraulic - those actuators are all pneumatic...

whoops:o

That's what I get for posting in a hurry...

Thanks for the correction Jon!

asbrandsson
01-29-2008, 08:46 AM
Ok to clarify!! On the mini and micro frontier servos are the way to go. But im talking about bigger stuff. Like say a crustcrawler that u can ride on! You guys do some really awesome stuff with the mini's and micro bots that I cant see why you cant transfer to a larger scale. I mean come on u have done all the programming, made a design, but why not go big. Aside from the cost of course i really see hydraulics being the only feasible option to make this work. For example http://www.neogentronyx.com/. I know 99% of you guys have seen this but this guy started backwards. No programming and no small scale design. For some of you the only challenge a project of this scale would be just enlarging your creations. Right now as far as i know this thing hasnt walked yet! But you guys could do something that can awe people. I know the post is kinda deviating from the topic but how would one use hydraulics for projects like this? I mean I just got into the whole scene of trying to build a robot this last year but i have researched this for a long time and i just curious why the guys that actually made some progress hasnt tried going big yet?
Hello,

I have seen some videos of that guys mech on youtube but I am not really sure that it works all that well as none of the video show all that much. And there is a guy inside of it controlling it. On a personal level I am trying to get something made that can contol itself.

Asbrandsson

hotrodled
01-30-2008, 12:03 PM
Check out the Hydraulics on this giant spider.

http://www.youtube.com/watch?v=kFZvXuOIVMs


http://www.youtube.com/watch?v=kFZvXuOIVMs

Too cool!

Alex
01-30-2008, 12:25 PM
ok, now I've seen everything! I wonder what in the hell would all of the neighbors think when they see someone trotting down the road in this!

kdwyer
01-30-2008, 01:28 PM
Smoooooth ride, dude!
This points out most of the pitfalls of hydraulics...
Power-hungry, needs an engine.
Overly complex (but I love all those linkages!)
Doesn't exactly have a 'tender touch'
OK, it didn't leak, but it will!

Besides all that, I think the kid on a trike had to wait for it to catch up. And turning...

All that aside, this is an amazing machine. I bet it worked fine as a model. One last big problem is that some things don't scale up (or down) very well.
As a proof-of-concept, it works. As a usable vehicle, it leaves alot to be desired. As a Hobbyist accomplishment, I take my hat off to him.

Now get some big batteries and electric motors.

DresnerRobotics
01-30-2008, 03:53 PM
I've seen that giant walker before... and I don't think it's actually hydraulic. I think it's basically just a differential driven crank shaft mechanism. I don't know what the exact technical term for it is, but I've seen some simple multi-legged walking robots that only have 2 DC motors that use the same setup. Basically a large version of this:

http://www.trossenrobotics.com/owi-spider-3.aspx

JonHylands
01-30-2008, 04:01 PM
I think you're both right. If you look carefully about 1 minute in, you can clearly see a couple hydraulic lines running out to the side. My guess is those lines drive a hydraulic motor, which is what provides the rotary motion for the crank drive.

- Jon

vcambri
01-31-2008, 10:11 AM
What if there was a system of expanding/contracting bladders that were connected to one central pump through a "nervous system" of lines where pressure in the bladders is controlled through valves in the lines. Each bladder has an in-flow and an out-flow line. Flow into the bladder is unimpeded. The state of a valve on the out-flow line controls pressure in the bladder and hence size of the bladder.
All "muscles" (bladders) would be controlled by a central processor through the small valves in the lines. The one large pump, like a heart in the center of the machine pumps fluid (call it blood) but here the "blood" controls the power state of the muscles not their food and oxygen levels.
Such a machine would come closer to the analog nature of natural beings. Motions would be more smooth and more like those of a human or animal.
I'm really interested in your thoughts about this...

MYKL
02-11-2008, 04:53 PM
Hello,

I've actually built models for hydraulic muscles that use bladders and sleeves a lot like the shadowgroups 'muscles' do. In my setup there is a positive displacement pump and two 'muscles' that oppose one another in the machine. Fluid volume is driven into the internal bladder causeing the 'muscle' to contract. This volume is pumped out of the 'muscle' on the opposite side so that when the pump is reversed the fluid actuates the 'muscles' in reverse and the limb moves in the opposite direction. Where I am at now is the prohibitively high price of a small efficient gear pump. I believe this system could make an acutuality out of a fairly light weight armoured suit with a waldo like manmachine interface alla Appleseeds Landmates. Or a combination of man-machine and CPU controlled movement as imagined in the Patlabor series of Anime.

asbrandsson
02-11-2008, 05:47 PM
Hello,

Not to scew up the thread, but it is good to see you MYKL over from Robosavvy. Why not put some stuff up about the hands you have been working on MYKL?

Asbrandsson

vcambri
02-12-2008, 07:37 AM
Hmmm. Interesting.
But the point I was making is that you do not need a reversable pump with adjustable output to actuate the limbs. You can have much smaller devices that control valves on lines feeding the "muscles". Then you can use one central pump with constant, unreversed output. Purportedly, that is...


Hello,

I've actually built models for hydraulic muscles that use bladders and sleeves a lot like the shadowgroups 'muscles' do. In my setup there is a positive displacement pump and two 'muscles' that oppose one another in the machine. Fluid volume is driven into the internal bladder causeing the 'muscle' to contract. This volume is pumped out of the 'muscle' on the opposite side so that when the pump is reversed the fluid actuates the 'muscles' in reverse and the limb moves in the opposite direction. Where I am at now is the prohibitively high price of a small efficient gear pump. I believe this system could make an acutuality out of a fairly light weight armoured suit with a waldo like manmachine interface alla Appleseeds Landmates. Or a combination of man-machine and CPU controlled movement as imagined in the Patlabor series of Anime.

MYKL
02-12-2008, 10:45 AM
I believe that the hydraulic system in an arachnoid works the way you describe. I tried this system first as it makes logical sense. Here are the problems I had with it:

1. You have to have a source and a return 'vein' running to each 'muscle'.

2. The only valves that are minute enough (and afordable) to work at the 'muscle' itself are actuated by electrical power and require cycling times because of cooling and restretching issues. (see NiTi based muscle wire valves).

3. You must account for a loaded actuator much differently than you would an unloaded one. This allows for some wild fulcuations in current and pressure demands. Your control of the valves must be VERY precise. In order to get the most out of a bladder type 'muscle' you have to have the power at the beginning of the stroke. The valve system delivers power in a ramped style as the valves restrict flow.

Using a servo driven power source be it Gear Pump or Cylinder is the next step in my particular development of this kind of actuation system.

I'm not trying to punch holes in what you are doing. If you can make it work, more power to you. I am not very proficient with the control electrics that can be programed to juggle the calcs it would take to facilitate a constant flow system. I'd love to see anything with these kind of 'muscles' make a statement in the robotics community.

I am very happy to find someone that is pursueing this goal. Someone I can bounce around ideas with.

Be well...

^_^

SN96
02-17-2008, 08:27 AM
Hydraulics require pumps, valves, regulators and they can leak. They do not provide smooth operation without a sophisticated control system. Because you cannot compress a liquid, there is no dampening.

Pneumatics provide a better solution for robotics because they act more like muscle by dampening movement, are strong, light, and less messy.

MYKL
02-18-2008, 10:48 AM
Therein lies the beauty of using the 'muscles' we've been discussing here. The actuator itself dampens the assembly. The fluid is moved by a PD pump. The whole system is closed, there are rod seals that can leak.

Robin Hewitt
02-29-2008, 02:07 PM
Hi Robo freaks, I'm a Newbie :veryhappy:

I spent a year or two walking around looking at my feet and trying to figure out how to do bipedal locomotion in a human sized frame. Did consider hydraulics using rolling diaphram pistons but it just didn't fit. The problem was not so much the pump or the pistons, it was the switching, couldn't control it without a vast bulk of solenoids and motors, no room. The only available space not full of musculature was the pelvic girdle and that was just too precious as a battery compartment, don't want that weight too high up. The chest cavity is full of spine bend, shoulder blade control and computers.

I eventually settled on Maxon rare earth motors powered by a stack of Cyclon lead acid accumulators in the pelvis. Did a lot of torque calculations, bought a lot of parts, drew a lot of AutoCAD, then everything stopped. The only way I could see to build it was laser cut alloy sheet and that was too constraining in the skinny bits.

Things have moved on over the past couple of years, I now have 2.5D CNC and can cut parts and injection tooling so the old bipedal ideas are starting to itch. I'm back to designing feet, strange things feet, the way they bend and everything knits together as they flex, I have spent a lot of time doing feet, best not start on feet and let's not even mention knees :eek:

I've probably rambled enough to prove my credentials, look forward to some interesting chat :veryhappy:

Robin

DresnerRobotics
02-29-2008, 02:15 PM
Hi Robo freaks, I'm a Newbie :veryhappy:

I spent a year or two walking around looking at my feet and trying to figure out how to do bipedal locomotion in a human sized frame. Did consider hydraulics using rolling diaphram pistons but it just didn't fit. The problem was not so much the pump or the pistons, it was the switching, couldn't control it without a vast bulk of solenoids and motors, no room. The only available space not full of musculature was the pelvic girdle and that was just too precious as a battery compartment, don't want that weight too high up. The chest cavity is full of spine bend, shoulder blade control and computers.

I eventually settled on Maxon rare earth motors powered by a stack of Cyclon lead acid accumulators in the pelvis. Did a lot of torque calculations, bought a lot of parts, drew a lot of AutoCAD, then everything stopped. The only way I could see to build it was laser cut alloy sheet and that was too constraining in the skinny bits.

Things have moved on over the past couple of years, I now have 2.5D CNC and can cut parts and injection tooling so the old bipedal ideas are starting to itch. I'm back to designing feet, strange things feet, the way they bend and everything knits together as they flex, I have spent a lot of time doing feet, best not start on feet and let's not even mention knees :eek:

I've probably rambled enough to prove my credentials, look forward to some interesting chat :veryhappy:

Robin


PICS!

Robin Hewitt
02-29-2008, 04:36 PM
PICS!

No robo-pics until I start cutting metal.

On topic, I have been doing some hydraulics recently. Here's a picture that shows the kind of mess you get into when you try to hold two different pressures on two rams using only one pump. Confidence being that feeling you get before etc. :eek:

triggerhapy
04-30-2010, 10:07 AM
maybe because you failed your grammar class?

defwheezer
04-30-2010, 11:16 AM
Sources of small hydraulics include makers of some the high end RC trucks like Wedico (http://www.wedico.de/eng/dumper.asp) and

Bernett
http://www.bernett-modellbau.de/

Damitz
http://www.damitz-modelltechnik.de/index3.htm

Leimbach
http://www.leimbach-modellbau.de/

Stahl
http://www.stahl-modellbau.com/

jdolecki
04-30-2010, 11:48 AM
I didnt know that there was a model market for hydraulics.

small pumps, cylinders and lines

This place seems the best for components

http://www.damitz-modelltechnik.de/index3.htm (http://www.damitz-modelltechnik.de/index3.htm)

wow

engineer
04-30-2010, 12:23 PM
for the total rc nut gardentrucking.com

tom_chang79
05-02-2010, 11:01 PM
Hi!

I don't now about humanoid hydralic driven bots. But I'm pretty sure this walking tractor use hydralics:

http://www.youtube.com/watch?v=CD2V8GFqk_Y

Very cool!

-Zenta

Will Trossen be carrying this tractor? I would like to buy this tractor after I save up enough money for it!

NovemberAlpha
05-14-2012, 08:16 PM
So Andrew...A "full-size" hydraulic unit... Funny, Mr. Mania and I are about to do just that. I'd love to pick your brain.

NovemberKiloEcho.com

Robin Hewitt
05-15-2012, 03:33 AM
I've been thinking long and hard about this... For humanoid I think take a small carbon fibre scuba tank, drop it to 10 bar, convert air pressure to liquid pressure then use servo motors with hydraulic assist to get the power. At low pressure you can use cheap pneumatic cylinders muscles. The tricky bit is the valves, but I think the answer lies in the sliding valve used by steam engines. You use the steam/oil pressure to hold the flat valve faces together. These valves bed in where the normal shuttle valves wear out.

Peter_heim
05-15-2012, 06:49 AM
Hi Robin
Hydraulics is about flow under pressure fluid from the pressure tank(accumulator) moves the actuator then returns to the tank via a pump to increase the pressure the cycle then repeats. The other way is servo boost the pressure takes the weight and then is dumped (best done with air)> At one stage i wanted to control a arm with hydraulics i have a few old actuators and hose the big problem i had was the valves my plan was to use PWM to turn both the supply and return on and off to lock the actuator in place and PWM to close the valve slowly to avoid hammer in the system. The main problems i had with the experiments was the noise of the pump and the leaks (from using old pipes) the power was great 1000 psi can really move a arm but the size was large to reduce the size requires money a 1/4 inch hose can range in price from $50 to $400 a meter with fittings. There are some good examples of hydraulic robots on Google and what they can do is great (look for the robot drummer) most have the sound turned off

Regards Peter

Robin Hewitt
05-16-2012, 01:14 AM
Hi Peter

Air is too springy and reversing direction means exhausting one side of the cylinder which is wasteful. If you drop to 10 bar you can use cheap nylon hose and push fittings. Trying to regulate the flow with a remote valve is practically impossible, moving the valve with the actuator to get hydraulic assist is self regulating and easy. If you replace the pump with a pneumatic cylinder having oil on one side of the piston and air on the other you only need to silence the exhaust.

best Robin

stevenjon918
11-03-2012, 10:49 PM
To difficult to make a useable pump for a sophisticated system would be my guess , what about useing an electrail charge to heat and exspand fuild in a micro line . If the fuild could be designed to act to a certin pulse or frequency then not only would you have not a need for a pump but you could just about make activity that coppied a human hand . The design of tiny receivers that could be set or programed, small enough to be mixed with fuild would be possable I would think some time in the near future , if not now . If even a transister , again small enough to be mixed with a fuild and placed on the inside of a matched tube , that would change how fast heat is develioped with in a closed system would almost be like an adrenalin surge to a micro hydraulic system .

Th232
11-03-2012, 11:12 PM
An interesting concept. However, the way I see it, the issue with a heat based system like that would be that of removing the heat. Either:

a) Heat can't leave the fluid very quickly. This will greatly slow down the response time of your system.
b) Heat does leave the fluid very quickly. This means you'll have to expend a lot of energy to keep the fluid heated.

The only way I'd see it being a viable alternative would be by having some kind of active cooling system or a means of cycling the fluid. Neither of those would be cheap/easy to do.

By the by, it's "fluid", not "fuild".:p

insert bill here
08-01-2013, 11:21 AM
They're showing up now in Google searches as, "miniature hydraulics." Someone in Germany is making amazing functional replicas of earth-moving equipment, using miniature hydraulics. Here are some links:

www.wpraez.de (http://www.wpraez.de)
YouTube video demo (Ohio, USA): http://youtu.be/WPvIpmOu_Zc
www.tobias-braeker.de (http://www.tobias-braeker.de)
YouTube videos of his models: http://youtu.be/8OeOa5oYaSU