Ok firstly Hi all I’m new
I’m currently planning a project to design a walking robot tripod that is also carrying a payload on the top. I have a picture of what the robot will look like though with my design I am planning on using carbon fibre tubing for the legs(4 to 6 legs).
Please have a quick look at this gallery to get an idea of the robot look
Cant post the link to the image as im new and its not allowed…so just look for “RoMeLa Robotics Media Lap at Virginia Tech” the Robot Im talking about is called STRiDER and stands for “Self-excited Tripedal Dynamic Experimental Robot”
What I’m trying to wrap my head around is how to take into account the torque I’m going to need for servos with such a long leg design (carbon fibre tube). Currently I’m looking at the Dynamixal AX-12 servo as it seems a great price and has many features. The payload I need for this is about 0.8kg or 1.8 lbs. This is rather allot but I can’t change the weight of what the tripod is carrying as I can’t modify the device. I would like it to walk the same way one of the hexapods does except much taller.
So if any of you have any ideas for my planning stage then please feel free to comment.
ok well the think you have is the right tobot anyway and I meant media lab not lap…just a typo lol. But basically thats the look of the robot i want to make, I like the whole walking fuction they have there but seeing as they dont get it to work that well yet Im not even going to begin trying. I would try a 3 leg walking robot but I can imagine that going to be pretty hard todo, and the hexapod movement seems fairly stable. Trying to find any good calculation methods for stress of the servos would interest me too.
One of the contest projects was using very long legs, and mentioned a Trossen Robotics tutorial on servo torque calculations. I’d check out his thread, and study the tutorial.
I’ve seen the strider before. Really need and unique bot! I think you need to dig deep in to dynamics because this isn’t easy (or even impossible) solved by IK alone.
On the servo torue, I doubt that the bot can get up by itself. Sinds the legs are so long you’ll need really huge servos to get it up. But the torque is rather small when the legs are vertical. I doubt the robotshop calulations will be of any help since the construction and purpose is different… (sry, don’t want to be a spoiler)
Try digging into IK and dynamics. The Strider looks very balanced; the secret behind this will be a lot of math. Maybe you want to build some simulation before starting on the mechanics.
thanks for the advice, ill have a chance to sit down and work that out sometime latter today, Ill post my findings as well. Ill proably model this in 3dsmax too to give a better idea of the end result/look. Then latter try to find some way of simulating the ik/walk/stand a sit cycle.
Interesting concept but I don’t see any benefit from this design other than the challenge of getting it to work, or the potential technology gained from achieving success. So far it looks like it only works in the cartoon world.
Long legs will definitely require high torque servos due to the distance from the axis of each joint. That quad stick walker you mentioned that was in the contest had a problem lifting its own body weight.
That design might have benefited greatly had helper springs been used on all the joints.
the look is because i use robotics as part ot my art practice…to get an idea maybe look at my website…there are a few tripod things there that do strange things
I looked at the referenced site, and the walking will have some unique challanges. The base servos will need to be set for continous rotation with some type of position feedback (encoders are mentioned). Slip rings will need to be used at the main joints where the continous rotation will occurr. During the step the bot will become dynamically unstable and depend on what I would consider a “lucky” balance condition. Definately a challanging project.
hmm. I really just want it to look like the strider not move like it does with the leg flipping under…I definatley dont have the equipment to machine parts for that…well not anymore. Just think of a normal hexapod that stands over a meter tall and has carbon fiber legs…so im sure it will be still much more challenging than a normal heaxpod because of balance but im sure if I have the right control and strong enough servos i can do it. But anyway I have to work on these calculations first then ill come back and post what I think sounds like the best fit.
Perhaps if you get it up on it’s feet first, then it won’t be so bad to move the legs. But those long legs will magnify any errors in positioning, while also “dividing” the torque!
Have you built a “normal” hexapod first? That might give you some experience in moving the legs. Theoretically the IK would be the same, as it’s a hexapod!
I suggest you purchase the servo you are counting on to do the heavy lifting and do load testing with it first. It will be the key part that has to work in your project. A chain is only as strong as the weakest link!
I’d have to agree with zoomkats suggestion that you buy a servo (or two) and mock up one or two of the worst case joints to make sure the servo isn’t going to just burn out.
actually im going to do this in the morning(im in Austria). I just had some inspiration a few moments ago because I was dreading having to look for a closer supply for the carbon fiber legs…then my flatmate came home after his archery practice…so Im off to buy 2 carbon fiber arrows tomrrow…plus 2 hitec servos from my local RC shop to test…the test will be in reverse seeing as i dont have more than one leg but if they can move the carbon fiber rods in the reverse direction then i imagine its a good start…
actually if your hobby shop has carbon fibre tape it can be used to laminate balsa spars. usually unwaxed dental floss is used to bind the tape to the balsa in a spiral wrap along the length and then it is saturated with thin CA. the result is extremely strong and about half the weight of a spruce spar. be careful handling the tape though because you can get splinters and they hurt.
ok maybe ill goto the hobbie shop before going to the archery shop and see how much that will be, I mean the arrows without feathers are exactly the right length and are only about $8.
