Moving this discussion to its own topic...

This forum has been going for a long time now and it seems like a lot of topics get recycled... particularly questions in one form or another about accuracy and distance. Questions may focus on pouch design, ammo shapes, throwing style, or the gravitational effects of the moon, but there’s no easy way to really definitively answer a lot of these questions by sharing opinions or anecdotes. Sometimes you need solid data to get a solid answer, but slinging is so complicated that controlling all the variables in the data is nearly impossible... especially since a human is involved.

Enter the Roboslinger. I hereby propose that we build an instrumented precision trebuchet, not to answer questions of human capability, but to maybe settle some of the questions of slinging mechanics by taking the human out of the equation. By precision, I mean tight tolerances and low friction components. By instrumented, I mean we actually put sensors on it to measure things like string tension and rate of rotation. And it would obviously have to allow you to attach almost any sling to the arm for apples-apples comparisons of slings. In other words, this is not your run-of-the-mill medieval siege engine. It’s a rock slinging science machine!

I am a robotics engineer by day, and I could probably just build something myself and tell you all about it, but after my family and side business,  I don’t have enough time left in the evenings and weekends to do this without some help. I’d rather make this a community project anyway. I have plenty of design ideas, but other people live in better places for testing it. Maybe it’s designed to ship around the world as needed?

Before we could start using a Roboslinger, we would have to come up with a design, a budget, build it, characterize it, and plan out what information we need to collect to answer what questions. This might also cost more money than I personally could put in, which means we may want to consider fundraising options like t-shirts, slings, or maybe giving away slings for GoFundMe donations.

What do you all think?

I'm all in of course. Anything I can do to help just let me know. I'm in the Houston area, not sure where you are at but if you are around here and need a second pair of hands, albeit unskilled, I will pitch in wherever and whenever possible.

As for design mechanics I don't know how complicated you are planning on going but I think one thing that is often overlooked is palm position in relation to forward momentum. That is something that interests me. There are many other things as well but plenty of time to discuss them later.

Thanks for being willing to do this. I think whether people immediately realize it or not this is a big deal for advancing slinging as a sport.

Sounds a great idea. I can't wait to see some data.

The simplest way to do it would be to use a 'rubber powered' trebuchet I would imagine. Joergsprave of the slingshot channel on youtube already did something similar.

https://www.youtube.com/watch?v=7vJBKfQFD8I

(starting around 11mins- I recommend the whole vid though, it's pretty interesting)

You could use your own throws to calibrate it to a realistic power level.

It would be pretty interesting to get a high speed camera on it to to see the speed or path through time too.

In terms of what I can contribute, I'm not really in a place to test nor do I have much resources to build. I'm also poor and busy  :P But if there's any maths or analysis I could pitch in depending on how much I've got on.

Morphy wrote on Apr 12^th, 2018 at 10:27am:

I'm all in of course. Anything I can do to help just let me know. I'm in the Houston area, not sure where you are at but if you are around here and need a second pair of hands, albeit unskilled, I will pitch in wherever and whenever possible. As for design mechanics I don't know how complicated you are planning on going but I think one thing that is often overlooked is palm position in relation to forward momentum. That is something that interests me. There are many other things as well but plenty of time to discuss them later. Thanks for being willing to do this. I think whether people immediately realize it or not this is a big deal for advancing slinging as a sport.

Thanks Morphy!

As far as how complicated... I would say it should be as simple as possible and only as complicated as necessary.  I doubt we will get everything right on the first try, but you can't learn what's not there until you have something to test out, so my preferred method of development is to build something, learn from it, then iterate.  The side-benefit is that there may end up being multiple versions of Roboslinger that serve different purposes and there would be more hardware to pass around.

As far as community engagement, I think that educators would be a really good target.  Whether it's history, anthropology, physics, biomechanics, aerodynamics, or mechanical engineering, there is something here for a lot of educators to use, and I could see possibly loaning the roboslinger out to academic researchers who want to use it to write academic publications as well.  The advantage of this approach is that, by reaching out to a few educators, those people can then reach hundreds or thousands of students who might take an interest in slinging generally as a result.

As far as location, I am in the San Antonio area a few hours away from you.  I also have some family in the Houston area.

JudoP wrote on Apr 12^th, 2018 at 11:00am:

Sounds a great idea. I can't wait to see some data. The simplest way to do it would be to use a 'rubber powered' trebuchet I would imagine. Joergsprave of the slingshot channel on youtube already did something similar. https://www.youtube.com/watch?v=7vJBKfQFD8I (starting around 11mins- I recommend the whole vid though, it's pretty interesting) You could use your own throws to calibrate it to a realistic power level. It would be pretty interesting to get a high speed camera on it to to see the speed or path through time too. In terms of what I can contribute, I'm not really in a place to test nor do I have much resources to build. I'm also poor and busy  :P But if there's any maths or analysis I could pitch in depending on how much I've got on.

That might be an option, but Joerg also shows in other videos how the force generated by the rubber depends on temperature, so that might affect consistency more than a gravity-based version.  With the right design though, I don't see any reason why you couldn't do either/or on the same machine... just swap out the weight for a rubber band as-needed.

An exciting project that I follow with interest!
But if it has a rubber powered, is it then slingshot-robo?
On the other hand, these are just concepts.

What about having it powered by something similar to a centrifuge, but with the ability to spin in multiple directions? Would that work? Might could even power it with something simple like a power drill. Or the power drill could be a temporary power source as the main body of the device is perfected, leaving a stronger source of power to be worked on after. I’m just throwing out ideas, hoping they may help in some way. As for the fundraising side, or anything else I’ll gladly help in any way I can. I agree that this, if it does come to fruition, could be a very, very, big deal.

Before we get too far down the road on specific designs, here are a few "ideals" I think we could aim for:

1. low cost - I could go either way on this one. It might be possible to find a rich uncle who is willing to spend thousands of dollars, but until then, I suspect that the project will be somewhat price sensitive.  Low Cost would also make it easier for other people to replicate the design.

2. Portable - I would like it to be small enough that it can be transported in the trunk of a car (maybe partially disassembled), and not so huge or heavy that it costs hundreds of dollars to ship it across the world.

3. Easy to use - Not everyone is a scientist or engineer, and it shouldn't take a PhD to use a rock chucker anyway.  Whatever electronics and sensors go on it should be either intuitive or invisible to the user.

4. Modular - We should be able to swap out parts easily with minimal tools.  It should be easy to attach any sling that is made for a person on there so we can test the same slings we throw with.  The arm should be swap-able so we can play around with stiff arms versus bendy things with simulated elbows and wrists, or whatever.  We should be able to put weights or rubber bands on the pulling end, or some other propulsion mechanism if desired.

5. high performance - this one might violate #1, but I think we can find a balance between the two.  Whenever we can afford it, we should use good quality components and keep friction on moving parts to a minimum... but there are also plenty of craftsmen here who can do quality fabrication work.  It doesn't necessarily have to be expensive to perform well if people are willing to put time into making components.  Particularly expensive components should become specific fundraising goals.

6. Community ownership - I don't want this to just belong to one person unless only one person does all the work and spends all the money.  Anyone who wants to use it or improve it and is willing to follow the community-defined rules should be allowed to do so.   Contributions of time and ideas are just as valuable as money too. Rules should be transparent and defined ahead of time so the roboslinger's use isn't ruled by the whims of one person's arbitrary decisions.  There should not be rules for rule's sake, but as problems arise, rules should follow to solve those problems as simply as possible with minimal burden on the community.  And anyone put in a position of trust as a decision maker should also have clear rules of accountability.

7. It should actually be useful for answering specific questions about slinging and for demonstrating different principles of slinging.

8. We should have fun!

That's the short list of design goals off the top of my head.  More goals are also welcome  :)

Very well said!!

I was thinking it might be possible to set up some type of community auction to raise money for the birth or the world's first roboslinger.

We have a varied and talented group here, perhaps members could donate things to be auctioned off and we could put that money towards the creation of this device?

Once a design is on the table and we have a concrete fundraising goal to work towards I would be willing to start us off by offering this for auction:

https://i.imgur.com/2UukkDT.jpg

This is only an image I found. Mine is virtually new. It's been used all of once to chop on a fallen log for about 5 minutes. If anyone else is interested in offering up purchased or handcrafted items please let us know so I can get a thread started after the design phase is complete.

Morphy, wow! That’s incredibly generous!  I might want to bid on that chopper myself! :D

Roboslinger makes me think of the golfing machine, Iron Byron. My old golf coach said I had a swing like the Iron Byron, though my shots still go all over the place  :P

stick me down for design and 3d printing.
Not a robotics engineer - but fairly advanced 3d printer owner and builder.

One way to make cheap and consistent parts is to simply print them when and where they are needed.
.stl files are pretty small and easy to email. much easier and cheaper than posting actual parts.

Sounds like a fun project :-)

As far as sensors and the like goes.
What platform you thinking of using ?
r-pi 3's are cheap and pretty comprehensive.
But plenty of other mini-system boards around now.

The size is going to be the trickeist part. Have to determine maximum length sling and arm length.
The rest of it's going to be fairly compact. But arm length is crucial.

Can I suggest aluminium 2020 extrusion for the arm. Lightweight, cheap, readily available and much more rigid than a tube.
Also a ridiculous amount of screws. bolts brackets etc readily available - and cheap. 

Oh yeah I'll sticky this thread.

Thanks CA!
I’m thinking about what you’re saying about extruded aluminum, and it brings up a broader point: Some people may be better at building with lumber, others can weld, and others may have a pile of extruded aluminum already sitting around the house.
At this point we don’t know what is best, but we also don’t care as long as it performs well, so maybe we aim for a performance specification instead of a specific design. Let people build it from whatever they want. If you can achieve the precision and repeatability with a trebuchet built out of Legos... why not?

If there is a large enough crowd involved, then we can make it more like a competition to see what design gets the best performance. I suspect that over time a good design will naturally emerge if people are competing.

In the short term, I think we should start with one Roboslinger (which means we still have to decide on materials), but that prototype can be used as a reference design to develop a spec and to develop the rules of a competition when we are ready to bring in a wider audience. A competition only helps with the performance part of the equation anyway, not the instrumentation or data piece of the problem, but it would be a great way to get lots of people involved and interested in slinging.

What about adapting a trap thrower?

https://www.youtube.com/watch?v=7BbUrWn7DZk

It might be easier (and cheaper) to adapt an existing machine than to design and build a a rig from scratch.

It's a good idea.

But by the time you'd modified it with electronics, longer arm - sling holder and sensors.
It's probably simpler and cheaper to start from scratch.

Does raise the question though, how are we powering the arm ?
Weights or - like the trap - a spring ?

For consistency the spring might be the better option. Also makes a more compact machine than a falling weight.

The basic mechanics of a trebuchet are pretty simple.
Far more interested in sensors - what, where and why ? :-)   

Might be best to start with a specific list of what you actually want to measure.

The "sword swinging robot" from Myth Busters keeps coming to mind the more I hear this being described. They used it again to try to curve a bullet, which is I guess from a movie. If you go to 2:25ish you can see the robot in action... though it doesn't tick off the portable, cheap, or even intuitive check boxes mentioned above, I thought I would share to show a robot that was making the somewhat necessary motion.

https://www.youtube.com/watch?v=5vZ4lCKv1ik

Also to note, like this robot, a use of pneumatics might be practical. I can envision using a bicycle pump to fill a small tank. Most people have access to a bike pump, or one can travel with Roboslinger.

FYI,
I registered roboslinger.org and roboslinger.com for use with the project.  Assuming I set it up correctly, then within 24 hours, those domains should start redirecting to this topic on the forum (depending on when DNS entries are refreshed and updated), so it will be easy to point other people to this discussion here and invite them to participate.  I'll look into some sort of free or low cost web hosting options to put project-specific information up (maybe just use Wordpress.com?) if and when we get to the point of needing more than a bulletin board.

I guess one of the open questions is: how important is the orientation of the swing arm when the sling is slung?  For some things, 100% vertical is probably ok, and a gravity-based trebuchet with weights would be the simplest solution.

For simulating a sideways or diagonal throwing style, something other than gravity (an air piston, hydraulic piston, rubber band, etc.) might be better, but whatever supplies the propulsion force will need to be characterized for consistency.

One idea would be to attach a cable where the weight would normally go and then place the band/piston/rocket engine/etc somewhere else so that you could pull from any direction through a pulley.  If you did it this way, you could turn the whole trebuchet sideways or tilt it at an angle as needed.

---

ok, roboslinger.com seems to be forwarding to this discussion ok.  I just fixed the .org, so it should start working soon.

Thinking about all the different things that need to be designed for roboslinger, here's a partial list.  Maybe people can start taking on individual components that interest them?

Base structure: This can be wood, extruded aluminum, welded pipes, or whatever, but it needs to be stable and rigid.  If it breaks down or folds, then it needs to reassemble and perform reliably when put back together.
It will also need room to put sensors, electronics, batteries etc.
I would also suggest that we mount bubble levels in a couple spots on the base, and maybe have adjustable feet to help with leveling.  Eventually, we might want to put the whole trebuchet on a sled that slides forward as it throws to simulate the forward step.  This would probably not be something to put into the first version though.

Propulsion: There are already lots of ideas and there seems to be a lot of excitement about this particular piece of the design.  It would be nice if we can swap out a big weight, rubber, or a cable that could be attached to an external power source (hydraulic/pneumatic piston or whatever)

Arm: Needs to be rigid and have a strong, reliable attach point for the propulsion mechanism.  We would probably want a rotary encoder on the arm to measure the swing rate.  We may also want some adjustability in the length of the arm and/or the leverage ratio between the propulsion side and the projectile side of the arm.

"hand": this would normally be a simple peg on most trebuchets, but we will probably want to have some options to simulate different finger spacing and maybe wrist rotation.  it needs to be relatively easy to attache a sling that was made for a person to use too.  We may want an electronic release mechanism that allows us to let go at either a specific position or at a specific time.  As far as sensors, we probably want to put at least an accelerometer in the "hand".  We may also want to measure the tension on the sling.

Trigger: By this I mean the mechanism that starts the throwing process, not the mechanism that releases the sling string for the throw.  We will want to pay particular attention to safety.  This needs to be reliable and be difficult/impossible to trigger by accident.  Bonus points if it can be operated from a distance or has some sort of lockout mechanism to prevent accidental release when the sling is being loaded.

Aiming mechanisms:  There will probably be several ways to adjust and aim it, but if it's a 100% vertical trebuchet, then the most annoying one will be left/right rotation. If we don't have some easy mechanism for making small adjustments left or right, then it will be hard to aim, especially on uneven ground.

My vote for Robosling version 0.1 would be a purely vertical trebuchet because of its simplicity. The mechanical design to simulate a sidearm throw would be significantly more complicated, and there are plenty of lessons we can learn with the sensors and propulsion mechanisms on a vertical roboslinger before we start tilting it sideways.

Pure vertical seems a good idea, I'm sure gravity wouldn't be too significant for the basic slinging mechanism anyway.

I'm probably not the guy for practical building stuff, really more of a theorist  :P (even if I was I can't have many tools in my rented place).

Happy to give some verbal input on design though.

To me vertical makes the most sense as well. Probably the majority of things we can test with a very simple design. Im going to focus on supporting the funding aspect of it. All my tools are in storage right now about 1500 miles away so as much as I would like to build the base it would be pretty hard to do.

If anyone has anything they would like to donate or auction please feel free to PM me.

Right now I have the Parang and I'm sure I can come up with more things if I get creative. I have some skimpy pics of me that should net a small fortune for the right buyer. Just FYI gentlemen. They will be going fast.

Morphy wrote on Apr 14^th, 2018 at 4:25pm:

I have some skimpy pics of me that should net a small fortune for the right buyer. Just FYI gentlemen. They will be going fast.

The thought of Morphy wearing a traditional Balearic thong will probably haunt me for a week!

I could drum up some things to donate too. If any serious money changes hands, then there would need to be a non-profit organization so nobody gets into tax troubles. I personally wouldn’t want the IRS or their overseas equivalents to think the Roboslinger fund was personal income and say someone owes taxes on it. I don’t know how that would work in Europe, but it could definitely be an issue in the US.

On an unrelated note, I just volunteered for a biomechanics study with a friend of mine.  In exchange for being a test subject for his study, he is going to let me try slinging while I am markered up inside of a motion capture system.  That means we will have some data on the precise motion of a human (me) during slinging.  My form may not be the ultimate example of stone-throwing perfection, but unless someone else has a $100k MoCap system sitting in their garage, my form is all you get :)   

The idea is that we will have some information about the 3D movement of a human that we can use to make the roboslinger move in more human-like ways eventually, or at least to be able to say "this is the difference between our roboslinger and how a person does it".

I will try to get several different types of throwing styles in if I can.  I'm sure I won't get everything that everyone would want, and I am even more sure that my form will be off especially for those styles I am not used to, but I'll do my best.

This is cool!
I’d there’s something I can do to help please let me know!
I’ll definitely throw in an item for auction too.

NooneOfConsequence wrote on Apr 14^th, 2018 at 10:46pm:

I could drum up some things to donate too. If any serious money changes hands, then there would need to be a non-profit organization so nobody gets into tax troubles. I personally wouldn’t want the IRS or their overseas equivalents to think the Roboslinger fund was personal income and say someone owes taxes on it. I don’t know how that would work in Europe, but it could definitely be an issue in the US.

You don't tell them :-)

The 'throw' will have to be vertical. Without a pre-throw windup, and I have no idea how to simulate that without using an actual robotic arm, you can't do a sidearm throw.

So there's an idea - instead of making a suped up treb - why not make a cheapo robotic arm mounted on an adjustable paltform ?

With the motion capture info, you could accurately simulate throws properly.

https://www.thingiverse.com/search?q=robotic+arm&sa=&dwh=945ad72c437276e

Plenty of designs out there.

The thing is that the vast majority of sling styles rely tottally on the windup to generate power and the only way to replicate that is to make an actual arm, mounted on a powered platform to simulate the waist.

Motors would be the main cost. But you can get some pretty beefy stepper motors. gears are cheap and fairly easy to make (from my point of view). I guess you could use standard arduino boards and gcode to run the throw.

Just makes more sense. 

At the end of the day, we need soemthing that can move pretty fast, but on the positive side we're not using anything heavy at the end of the arm.  

So let's make an actual simulated slinger, rather than something static like a trebuchet that can only replicate the release part of the throw :-)

You need 4 joints: waist (rotating platform) shoulder, elbow and wrist.
You can use 2020 aluminium strut for 'bones'. Should be a way to make sliding joints to vary effective 'bone' length.
The mechanics are fairly straightforward. It's the software that's the bugger.

I'm curious, have you already thought of experiments to conduct?  I would like to hear some examples.
what variables will be eliminated by the robot?

Hello.
I'm new to slinging.org. I used to sling as a kid in the
70's-80's and picked it up again a couple of years ago. I
mostly enjoy throwing for distance. Does anyone have
any videos showing or demonstrating sling styles for distance.
thank you.

just use a centrifuge....shouldn't be too hard to fix up a throwing system to it.  u could control more variables that way=speed, acceleration, ect..

Sarosh wrote on Apr 18^th, 2018 at 12:26pm:

I'm curious, have you already thought of experiments to conduct?  I would like to hear some examples. what variables will be eliminated by the robot?

I would love to hear ideas. I personally want to do some accuracy experiments with a precision timed release mechanism and do some apples-apples comparisons of different sling designs. A vertical trebuchet would work for those experiments but CA is right about not being able to replicate sidearm throws on a treb.

@CA: a 4-DOF robot arm is pretty hard to build correctly on a budget.   In particular, the backlash on budget servos and gear boxes would make it hard to get repeatable results, and the system would be heavy too compared to a treb. It’s probably better to buy a $20k industrial arm rather than build my own if that was the way we wanted to go.

rcoon8865 wrote on Apr 18^th, 2018 at 1:10pm:

Hello. I'm new to slinging.org. I used to sling as a kid in the 70's-80's and picked it up again a couple of years ago. I mostly enjoy throwing for distance. Does anyone have any videos showing or demonstrating sling styles for distance. thank you.

Hi coon, there are quite a few threads on distance throwing and different styles on this forum. I would recommend looking at the FAQ and maybe doing a keyword search. The roboslinger topic isn’t the best place to discuss your question.

Quote:

It’s probably better to buy a $20k industrial arm

that confirms it - you are made of money :-)

At the end of the day it does depend what you actually want to test.

With a treb you can only really test forces at release, and effect of release points and methods.

pretty limited.

Don't see the problem with a robotic arm, plus it would be way more fun to build :-)

Curious Aardvark wrote on Apr 19^th, 2018 at 6:27am:

Quote: It’s probably better to buy a $20k industrial arm that confirms it - you are made of money :-) At the end of the day it does depend what you actually want to test. With a treb you can only really test forces at release, and effect of release points and methods. pretty limited. Don't see the problem with a robotic arm, plus it would be way more fun to build :-)

Ha! Not quite made of money. If I were, I wouldn’t need to be resourceful :)
It might be fun to build one, but it would not satisfy the current project. My point is that industrial arms are so hard to build from scratch that drumming up $20k would probably be easier. That definitely doesn’t mean it’s easy to just buy a robot (unless Morphy has a few gold plated emerald encrusted parangs to sell too).
Any robot arm that stands a chance at the performance we would need would at least cost thousands in parts cost alone. You can buy some robotic arms in the $10-20k range that would perform better and wouldn’t take years to finish building. Even those are not likely to give the performance we would want.

The other problem with electric motors is that they perform best under one specific load. That’s why you need to change gears to get a wider range of performance from a motor, but I’ve never seen a robotic arm with variable transmissions at the joints. Electric motors are a terrible surrogate for a human arm which can activate more or fewer muscle fibers at any time and transition smoothly between high strength and high speed motions while maintaining precision control the whole time. 

I would love to be proven wrong. It may be possible to get the performance we want on a budget, but I am not qualified to do that myself... at least not by building a robotic arm.

On a related note, Boston Dynamics did some impressive throwing experiments with an arm on a 4-legged robot, but it was using its whole body for the throw and the robot cost millions of dollars:

https://m.youtube.com/watch?v=-q7RsdxE76k

If we are looking at vertical throws we would most likely be replicating the Apache throw. Might I suggest you include that in the slow motion capture?

The thing with the figure 8 is the sling isn't just traveling vertical. It's coming from the opposite shoulder side from the sling. So there is some sideways momentum in play.

I did the capture yesterday. I’m not great at Apache, but I did try it. I will try to post some animations as soon as I get the data back. Might take a week or two for me to have anything to show.

The vertical-only design is fairly limiting, which is why I only said that should be our first iteration. We can always make it more complicated later, but a vertical treb will help get us started more quickly. One thing I’ve been thinking about is putting the sling on the outside of the trebuchet struts instead of between two vertical supports. That way we could add a side-swinging component to the trebuchet arm later on without a structural redesign.   I’ll post a concept sketch when I get a chance.

FYI, I did three sets of four styles in the mocap:
Underhand, Apache-ish, sidearm, and figure-8

Once I get the animations together, you can all debate whether I put the right labels on the four styles I used and whether the form was good or bad :)

that'll be very interesting to see.

You could add 'things' to the end of the treb arm that could simulate different arm movements.

Kind of like rubber powered, joints that just fire with the main trigger.

Could get some interesting and consistent swings like that.

I'm definitely not talking industrial robot arms. Or the things on thingiverse really as they're all designed the same way as a conventional arm.

One designed specifically as a slinger would be quite different.
I haven't seen any home made robot arms that use metal 'bones'.

Give me a while to think about it and I could easily design the physical side of it. Just don't know the first thing about programming and electronics.

I guarentee it wouldn't look pretty. But I come from the ducktape and string school of building stuff.
If it works - it doesn't have to be a work or art :-)
   

I discovered slinging as a distraction from my trebuchet experiments. You may want to look at Leonard Vance's projects on YT especially the "Murlin" trebuchet. A five arm highly efficient machine. You may find that significant research has already been done in this area by a number of individuals and educational centres.

Al Paca wrote on Apr 23^rd, 2018 at 3:36pm:

I discovered slinging as a distraction from my trebuchet experiments. You may want to look at Leonard Vance's projects on YT especially the "Murlin" trebuchet. A five arm highly efficient machine. You may find that significant research has already been done in this area by a number of individuals and educational centres.

Thanks AP!  I’ll check that out.

Very interesting treb vid! Thanks for the tip Al Paca!  It definitely gives me some ideas for simulating a human throw too.

For reference, here’s the most informative video on the murlin:

https://youtu.be/K7mB5iftarQ

Thanks very much. Great vid

Very impressive but how do we apply this to slinging?

lol was thinking that.

On the other hand it's a really neat compact throwing device.
Getting a trebuchet - of any type - to simulate a sling throw is always going to be really difficult.

This looks easier to adjust than a standard treb, and easier to build, transport and store.
Plus the windup for the actual sling is far more like a slinging windup than a standard trebuchet.

Very very interesting. 

Mersa wrote on Apr 24^th, 2018 at 5:02am:

Very impressive but how do we apply this to slinging?

... you’ll see  ;)

That multi-arm cam system on the murlin allows you to tune the acceleration profile of the sling to some degree. A slinger can feel the stretchiness of the sling as it swings around and can respond in subtle ways a treb can’t, but a Murlin style treb is more tunable than a standard trebuchet.

The goal is not to exactly replicate human throwing. Instead we really want to use a simplified but consistent thrower to study slinging and improve both human throwing techniques and sling designs.

A treb lets you study the effects of sling stretch, pouch geometry, release timing, projectile spin, and so on. These are things most trebuchet builders don’t pay attention to because they are focused on maximizing distance only.

NooneOfConsequence wrote on Apr 24^th, 2018 at 9:08am:

sling stretch, pouch geometry, release timing, projectile spin, and so on.

i'm really interested in sling stretch and spin orientation! :D

but with such devices we can't learn a lot about what movement to perform to impart the most energy into the projectile (move our body efficiently) and what technique is better suited to different body types. :(

Sarosh wrote on Apr 24^th, 2018 at 11:48am:

NooneOfConsequence wrote on Apr 24th, 2018 at 9:08am: sling stretch, pouch geometry, release timing, projectile spin, and so on. i'm really interested in sling stretch and spin orientation! :D but with such devices we can't learn a lot about what movement to perform to impart the most energy into the projecthttp://slinging.org/forum/yabbfiles/Templates/Forum/default/right.gifile (move our body efficiently) and what technique is better suited to different body types. :(

I think there will be something for all of us to learn, but it certainly won't eliminate the need to practice.  That would take all the fun out of it anyway :)

There is plenty of info about optimizing energy transfer on the physics side, and while every body is different, muscles and nerves do behave in similar ways for most people.  There may still be things you can do with techniques or sling design to dial in your own throwing performance based on problems that you might observe when compared to the roboslinger's behavior under various conditions, but to get detailed information about a specific person, you'd probably have to use a motion capture system and do a lot of analysis.

Ultimately, we'll just have to build the darn thing and start slinging to see what we can really learn :D

To that end, it seems that a rough design is beginning to emerge.  Maybe Curious Ardvark or someone else with some mad CAD skills could draw up a prototype small-scale murlin treb for the 3D printer as a starting point???

I meant to put this up a while ago... here's an over-simplified structural concept that supports the treb arm only from one side.  Add in a modified murlin pulley and cam system instead of a standard arm, then add a trigger, and it might start to look like a design instead of a cheesy MS Paint sketch, but for now you'll have to use some imagination.

The whole point of this sketch is that we could add a lateral degree of freedom (i.e. elbow) later on if we only support the trebuchet arm from one side.

---

Very interesting thread; I haven't been on the forum for a while so I've been catching up. One idea that just occurred to me would be to attach a sling to a clay pigeon thrower (trap). It probably would have to be a small sling but it would be an inexpensive way to get a mechanical "arm". I've got one around somewhere, so maybe I'll try it out, but this time of year farm work takes priority over any sling experiments.

-Timothy Potter

Timothy Potter wrote on Apr 25^th, 2018 at 3:51pm:

Very interesting thread; I haven't been on the forum for a while so I've been catching up. One idea that just occurred to me would be to attach a sling to a clay pigeon thrower (trap). It probably would have to be a small sling but it would be an inexpensive way to get a mechanical "arm". I've got one around somewhere, so maybe I'll try it out, but this time of year farm work takes priority over any sling experiments. -Timothy Potter

If you do, please post a video. I’d love to see it in action. Even if it doesn’t work perfectly, I might need a good excuse to buy a skeet thrower anyway :D

Well, it was a little more complicated than I expected to make these animations, but after processing the mocap data through 3 different programs (one of which I had to write myself), here are the results...

I can reprocess the same info from any perspective to produce new animations, and I have data on the speed and acceleration of the sling throughout the throws. Unfortunately we could not put a marker on the ball I was throwing.

The projectile was a 23g plastic wiffle ball, which was too light, but it was all I could throw around all the expensive equipment  :)

---

These are cool.
What were the pouch speeds like

Mersa wrote on Apr 27^th, 2018 at 6:09pm:

These are cool. What were the pouch speeds like

I have only had time to process data from one of the throws so far (out of 12), and I’m not at my computer right now, but I will show some graphs of hand speed vs pouch speed and pouch speed at release, etc. over the next few days as I have time.

Awesome!

wow :D
who needs a roboslinger :P

That's really cool. Man, I love technology.

I agree, that's amazing. Is it possible to be able to see the amount of tension on the cords at any given moment? I'm guessing that's a no, but that would be really interesting to know.

Morphy wrote on Apr 28^th, 2018 at 10:04am:

I agree, that's amazing. Is it possible to be able to see the amount of tension on the cords at any given moment? I'm guessing that's a no, but that would be really interesting to know.

If you could get a read of the acceleration of the projectile throughout the stroke you could probably work it out.

Morphy wrote on Apr 28^th, 2018 at 10:04am:

I agree, that's amazing. Is it possible to be able to see the amount of tension on the cords at any given moment? I'm guessing that's a no, but that would be really interesting to know.

I have a very accurate weight for the sling and pouch. I t would be possible to calculate the tension on both strings since we know both the mass and acceleration, but we wouldn’t know if the load is distributed unevenly between them.

@all:
I know I’ve been quiet lately on Roboslinger, but I should have a few things to show off again soon. Since the move to maths, I tested the roboslinger.org and the .com links, and they still point to the right place.

Has Roboslinger been issued an Individual Score Book For The Sling?

Have they been keeping detailed records of their throws?

Can we use a similar approach to model ammunition?

Can we put it all together and model variance, possible/probable group-sizes (Cone of Fire) and so forth?

I'm a Luddite, so I have no idea how hard it would be to program that model. Please educate this ignoramus....

At this point, Roboslinger is just a concept waiting to be designed and built, but the whole idea is that it should be a data-generating beast that gives us high-quality results by taking the human out of the loop and letting us characterize the sling or the ammo in a repeatable way that is similar-ish to how a human would actually throw.

Posting this raw mocap data for Montanan Slinger... anyone else is also welcome to use these files for personal use, but please talk to me before using it for any commercial use.  As a courtesy, I mostly just want to know who is using the data and what for, but if you animate a slinging character in a videogame and make millions off of my mocap data, you might at least owe me lunch ;D

These files can be viewed using a free program called Mokka:
https://biomechanical-toolkit.github.io/mokka/index.html

---

For those with enough money:

https://www.youtube.com/watch?v=LAtdsDTt__s

Just throwing it out there  ;).

Nice!  It doesn’t quite meet the criteria, but that does look like fun!

It’d be even better if it threw more than 100’... but to be fair, I can’t quite throw a bowling ball that far when I go slinging  :D

FYI, I bought a high speed camera this week!
Well... sort of high speed. It’ll do very low-res at 1000fps, but it was the best $130 used Casio high speed camera EBay had to offer!

It should come in handy for studying sling mechanics anyway. I will post some vids after it arrives.

https://www.youtube.com/watch?v=ND8qfS8gkDQ

https://www.youtube.com/watch?v=-gn2RGPqe_A

https://www.youtube.com/watch?v=KpFTyE-wiNo

https://www.youtube.com/watch?v=mXhLdz-Qmsk

NooneOfConsequence wrote on Nov 17^th, 2019 at 9:49am:

https://www.youtube.com/watch?v=ND8qfS8gkDQ https://www.youtube.com/watch?v=-gn2RGPqe_A https://www.youtube.com/watch?v=KpFTyE-wiNo https://www.youtube.com/watch?v=mXhLdz-Qmsk

great videos I think I've shared them already somewhere in the forum.
what I've been wondering is : if we got a roboslinger or a trebuchet or a program showing us the optimal trajectory of the projectile before release , would we be able to transform those ideas or images into movement? would those information be of any use to us?
e.g. what did we learn from the above videos that will improve our slinging technique?

video analysis of oneself is much more useful.

there is such a program on here somewhere - one of matthias's little works of genius.

One of these days I need to go through the backup and find all this stuff.

Sarosh wrote on Nov 17^th, 2019 at 12:35pm:

NooneOfConsequence wrote on Nov 17th, 2019 at 9:49am: https://www.youtube.com/watch?v=ND8qfS8gkDQ https://www.youtube.com/watch?v=-gn2RGPqe_A https://www.youtube.com/watch?v=KpFTyE-wiNo https://www.youtube.com/watch?v=mXhLdz-Qmsk great videos I think I've shared them already somewhere in the forum. what I've been wondering is : if we got a roboslinger or a trebuchet or a program showing us the optimal trajectory of the projectile before release , would we be able to transform those ideas or images into movement? would those information be of any use to us? e.g. what did we learn from the above videos that will improve our slinging technique? video analysis of oneself is much more useful.

There are a few things I want to use a treb for. One is sling pouch studies. If you have a consistent throw and the same ammo, you can analyze the difference in release timing from different pouch shapes.

You can also study ammo ballistics by using a the same pouch and varying the ammo geometry.

There’s two examples of how a treb can contribute to slinging. You can’t get that information from video analysis of people except in a statistical way.

NooneOfConsequence wrote on Nov 20^th, 2019 at 6:04pm:

There are a few things I want to use a treb for. One is sling pouch studies. If you have a consistent throw and the same ammo, you can analyze the difference in release timing from different pouch shapes. You can also study ammo ballistics by using a the same pouch and varying the ammo geometry. There’s two examples of how a treb can contribute to slinging. You can’t get that information from video analysis of people except in a statistical way.

ah ok I was thinking about the internal ballistics of slinging , "how can we develop our technique by watching a trebuchet?"