1webby1
General Category => General stuff => Topic started by: webby2 on April 04, 2020, 06:02:10 am
-
It has been about a year since I had to stop playing with my stuff and I am finally getting back to playing with things.
What I was trying to do is simple,, I have been trying to pass an input into a flywheel that is mounted on the end of an arm that is free to move. This input is from a source that is mounted on the very arm that the flywheel is mounted to and the goal is for me to spin up the flywheel without imparting a rotation into the arm as well.
Some of my prior printed test-bed parts made the move but not all of them,, some tests I need to redo since it has been so long since I did them,, and all that kind of stuff that happens when time passes.
I am not in any kind of hurry so I will take my time and post things when I feel like it :)
-
with what I have, as far as parts go and a few that I printed out I have started running a few basic tests just to bring myself back up to speed with where I was.
I purposely built this setup a little loose, that is it was supposed deflect a little bit to provide for a visual indicator of where the forces are heading and stuff,, well it does not flex so much,, so much for that idea :)
-
Fun stuff so far,, but nothing that I can say is conclusive.
Using spring force gauges for things that are not constant does not make for very good measurements, they are more of an indicator without giving exact numbers.
The indications then are that I can indeed accelerate a flywheel mass without the associated forces imparted into the arm that is holding it , what I mean by indicates,, well the force meters appear to show the same change in forces and in the directions I thought that they should,, but appearing to show and what they actually are are two different things.
I am building another test-bed,, this one is smaller :) I find that funny since almost everyone wants to go bigger because bigger will show it better!!! I am also trying to make it so I can use the same input direction and spin the flywheel in either direction,, a choice by design then can be made as to which way the flywheel spins.
-
Not sure which one to post this to,,
I have my printer printing some small gears,, I never was able to go below a 0.75 module before, but with this printer I am down at 0.5M,, it is kind of nice.
-
I had the basic mini system all put together and tried using the mini motor to run it,, unfortunately the mini motor is not strong enough to overcome the losses due to my printed gears,, I will try and find some others that will work,, the motors may have to grow a little bit but I am still going to try and keep it pretty small.
I could get it to spin up if I used 2 mini motors tagged in-line and wired in parallel but that is not going to work for me as a solution.
-
I have 2 builds that I am working with right now, I have the mini-build and I have my "other build".
The other build is kind of modular, that is I have made it so that I can take it apart and re-assemble it in a few different configurations which is nice in that it only takes me a few minutes to take it apart and put it back together again. I had to make a compromise with this build, not wanting to spend the money for some parts to do it better until I could see if it would be worth the expense. I have been playing with it while waiting for parts to show up for the mini-build and it has been interesting and it appears like I can use it to end up with a flywheel mass on an arm being accelerated without a net influence to the arm,, but due to the compromise I made I can not say for sure but what I see is enough to warrant spending the money,, more wait time for those parts to show up.
The mini-build is getting close, I have this pesky gear to get printed and then I move on with the other parts. I am not building this one with all of the controls in place, that is above my pay grade, but I am going to try and just balance forces the best I can. It will work in a sort of DC offset AC pattern kind of way with a small hint of my simple mechanical rectifier at work. My first test-bed for this one showed a lot of promise but the compromise I made for that one precludes it from showing anything absolute except that the forces are in the general place and direction I thought they would be,, just not able to rule out any other forces that I have not thought of or seen.
-
Interesting but most likely not useful,,
My other build is very modular, meaning that I can easily take it apart and then re-assemble with the parts in a different configuration so that I can change the path all the forces take to get to the same end point.
I have it together in one orientation and did not have any external mechanical connections made and was checking on the internal added force stuff,, well I fired up the internal added force and the system started to do its thing and then I noticed that being free to spin about its axle the whole system started to rotate and the mass of the system was also rotating and it was rotating in the direction of rotation of the whole system. The rotation of the whole system seems to settled down at the same rate of the internal rotation.
I reversed the polarity of the internal source and the rotation also reversed. It takes a little bit of time for the whole thing to stabilize but once it gets itself organized it settles down quickly to the stable RPM.
I think it is interesting reaction.
-
my modular build is not as modular as I would like,, normal stuff there,, but while I was working out some other parts I would like I think I ran across an interesting situation.
I think I have a twofer,, for those that do not understand that word it is short for " two for one".
It is elctro-mechanical in design and I made a few parts to put a quick test-bed together to see if what I thought could be there is there, in some form or another. That crude play time seemed to support what I think I can do so I am now making and designing some parts to do a comparison test since the electrical stuff I am using is not very efficient I can only compare these significantly less than 50% efficient devices input costs against the same load without my system and going through my system.
I am doomed to fail of-course,, but it is fun and exciting anyway.
-
I am using motors like these, as an fyi,,
https://www.amazon.com/Cylewet-Motor-Shaft-Arduino-CYT1037/dp/B01N9MS3UZ
They suck as far as efficiency goes :) They do work great for building motorized toys,, a few cars and tanks and well I had them lying around so I used them.
-
While I am contemplating the next step and changes to twofer I am working on 2 other test-beds that I am using to try and look closer at some of the relationships I am using in twofer.
I am planning on changing the drive system for twofer, I am going to move it external of the system to make it easier.
I also need to balance the system, this seems to have a very large impact on the draw of the input motor.
I am NOT planning on actually pulling twofer completely apart, but instead will try and use add-on parts only,, keeping the first test-bed intact could be a good thing.
-
So I am actually working on 3 test-beds as well as re-doing a bunch of parts for twofer,,
I only have 1 slow printer,, but I am not is a big hurry.
-
Strange thought.
If this system is working the way I think it is then there might be a way to utilize the system to provide for a force that can oppose gravity.
-
Back to redesigning things again.
What I think I am playing with is a torque limited device, that is to say that so long as I can supply a torque that is greater than the resistance the system will accelerate regardless of how fast the source of torque is rotating.
I am designing and printing out new parts with what I think is a reasonable setup to get there. With one test-bed I ran into a lot of air resistance, which is a torque to overcome, that limited the RPM of the system. I am going to run into that condition with any test-bed but if I know what that cost of rotation is then I can supply enough input to overcome that and the system should then keep accelerating. I know that this sounds just like a normal motor but in this case the RPM of the motor should not change I would just need to have the motor pass more current at the same RPM which is accomplished by providing a higher voltage to the motor, again sounds familiar but the RPM of the motor should not change.
This is what I am going to be building for and testing for.
-
still working on it but one test run I have made spins up faster than it should, at a cost of torque, but it is still limited in the amount of RPM it reaches.
-
Funny thing,, When you apply a force between 2 non equal masses usually the larger mass does not move as much as the smaller mass, in this case the larger mass is trying to do all the moving :)
Not sure what that means, it could be a frictional loss thing or maybe something else.
-
first,,
I cleaned the test-bed up and balanced it and used a better bearing setup and this condition stopped, it now is as would be expected where the larger mass changes much less than the smaller mass.
second,,
I now can accelerate a mass without influencing the arm it is sitting on very much, then I can decelerate that mass and have that influence the arm a lot, as in extracting back out of the mass all the energy I put into it to accelerate it,, minus frictional losses of course.
I am going to try a few changes to the existing test-bed to see if I can improve what it is doing and try and set it up so that it is completely autonomous.
-
Let me be a little clearer on this new thing.
I do not believe I can recover directly form the motor I am using as an input device to raise the potential of the mass the amount of energy I put into the motor.
What I think my test-bed is showing is that throwing away some of the input energy is allowing me the possibility to extract work in a different method, or path, than the work used to raise the potential of the mass.
I am going to make a change that removes the loss mechanism, this should lead my test-bed then to extract some work while the mass is slowly accelerating up to the point where I can no longer accelerate the mass and the system will become a constant state.
Then I am going to change the amount of potential I throw away to try and find what would be the optimum relationship for my test-bed.
-
I was having some interesting results,, but I found the "hidden" influence that I was missing.
I use gravity often a a source of resistance and when you have two different rates of acceleration and don't think about it correctly what looks like a non reaction just might be that the rate of acceleration is so low that the small amount of gravity force being used is enough to stop things from changing,, then when the rate of acceleration is much higher you get a change.
What I was missing is the very simple basic fact that I was applying a small acceleration against the same mass for a longer time than the reverse amount of acceleration over a shorter time period,, f=ma so there was a much larger force for shorter time.
-
I think I have identified 2 fatal flaws with my test-bed and the changes I made to it.
I am going to address those flaws and see if the changes make a difference.
My mass distribution is way out of wack with what it is I am converting, as well as the current chosen method to do that conversion.
-
Breaking parts is part of the game,, I break a lot of things.
What is annoying to me is when I make the same mistake repeatably. I try and keep clearances close but usually large enough to accommodate what I am doing,, but here I am having to re-do a bunch of parts because, like I have done many times before, I forgot to take into account the amount of shaft deflection there could be when that shaft is loaded,, so the 1mm of clearance I built to is not enough and things hit.
-
Some days when you keep making the same stupid observation and ignoring it, it is really silly after all, maybe you should go ahead and try and utilize it and see if there is anything there.
This is what I am doing now, well right now I am printing off a big part so when that is done I will start printing the parts to "test" my stupid observation.
If this works,, well then to make more out for the same RPM I would only need to build it bigger.
Simple concept is that the stronger force will overcome the weaker force leaving some force left over.
-
Things are very interesting.
I am going back, way back in time as to how things were done, and then changing it up so i can use an electric motor as a load device/drive device for another system.
My view on torque is that it is a closed system, otherwise you have leverage and not torque.
-
This is my desk with a small sample of the test-bed parts I have made.
Within this small sample are some of the parts I used to "test" my concept of how to create and apply a torque.
My solution is but one of many ways to go about it and now I am going to design a full system to run a real test.
-
I keep teasing myself with this build...
I printed off a few quick parts to make a slightly better finger toy version just to make a better decision on things,, Well that finger toy, being run by Mr. Hand, seems to show exactly what I thought and not what I was afraid would somehow be there.
Now to start with several very long prints,,,
-
funny thing,,,,,
As I am sitting here making the rest of the parts, I have the main "work horse" part built, a simple test came to mind to see if what I think I am doing is actually what I am getting and,, well,,, it is :)
This is the "work horse" part,, there are many parts inside it.
-
most days I am my biggest distraction and obstruction.
So I have been playing with the work horse and then for some reason I decided to change it and printed out another setup,,
It started out to just downsize the test-bed so that it could all be printed on an average 220mmX220mm printer, mine is larger at 310X310, but then while doing that, for some reason, I decided to add more complexity into the system and got all sidetracked with that instead of going back and running some very basic force measurements on the work horse. I think I decided that what I think I have is in fact what I have without actually testing it all,, silly behavior on my part.
With all that, today I ran few basic force tests on the work horse and,, well,, I did in fact get what I thought I would for the most part, but one reactionary force that I was planning on being 3/4 of the input\transfer force seems to be much lower and may be able to be almost canceled,, almost,, meaning most of my input could be converted into my output without a lot of reactionary forces interfering :)
-
still at it :)
I have destroyed a few of those yellow motors in testing some setups of what I am doing,, I showed one of the finger toys to my Son and he worked it and was puzzled over the "how" it could show him what it does.
Using all of this stuff to hopefully chase down that mysterious force of opposition, not the equal and opposite part,, is actually kind of fun :)
-
Feeling rather frustrated as of late.
You know how it goes, you build something and test it, make observations and then try and draw a few conclusions that you can then modify the test-bed for and see if what you think will happen does.
I have gone through a few mods using the workhorse setup and things keep going in the general direction that I am predicting they should, forces and directions working this way and that as I think they should.
Now I have taken all those observations and am trying to mod the test-bed to use them but unfortunately it might be that the amount of force I am throwing away, or that I need to throw away, is more than my little motors can handle. I am also failing at making some smooth guide pieces and so parts hang up on them,,
I am going to take a breather and think things over,, I might just have to design and print a completely new test-bed in order to incorporate the interactions I want to.
Lets just say that my little motors can make 10 N-cm of torque and I need to throw away half of that, so I am left with 5 N-cm of torque, now if the system is out 16cm from the pivot that will leave me with 0.3125N of force,, the reset device can use that up quickly.
The motors make more than that but I also need to throw away more than half,, well I should say that I am not using more than half rather than say throwing it away.
Either way I am barley making enough with all of the roughness and bad design things going on
-
Well I have a workaround for printing stuff and have almost all my parts printed for my next test-bed,, this one has taken a while.
I am hopeful that it will show me more interesting things, even if it fails it should provide some knowledge and that is a good thing :)
-
interesting first run-up of the test-bed.
I have to order some small parts before I can really start running some tests but with what I have I have managed to cobble enough together to test the motions,, basically.
What I have so far is that when I take a load out of the system by a generator the reaction the system makes is in one direction, but when I use a mass the reaction is in the other direction, this is not what I was expecting.
The generator I am using is not so great and can not make much resistance at the rate I am running it, with that, the difference in reaction force between the generator and the mass is huge, the mass creates a much higher reactionary force, until of course it is up to the system speed.
Running the system with no output load does what I was hoping it would, which is nothing much of anything except the reset sub-system does create a load on the drive motor.
An interesting day :)
-
If, and this is a big if, the system is working how I think it is then I might need to reconsider something I have just assumed for a long time and that is why an arm with a motor and flywheel accelerates the way it does.
If then I ponder that with how my system is reacting, and change my assumption of forces to being more like a differentiated force reaction as seen by the points of observation, then the behavior is appropriate, or it could be a form of precession and my assumption still holds.
I am not set up to overly test the position of the mass, I built the system to use a generator for the output load, I just placed a mass on the end of the output shaft.
-
I think I have most of the main system in the correct configuration, and I have those parts all printed and stuff.
Now for the more challenging part, the part that gets both frustrating and exciting :)
I now know what I need to do and where it needs to happen, and I think I know why,, it is a simple thing that may not be as simple to actually design and build.
When I get that figured out and added to the system then hopefully it will do what I would like it to do.
-
I designed and built my system with a part in the wrong place,, I am redoing it and moving the part to the correct place :)
Funny thought, what if I were to pull on something and it moved further away from me? Well that is what I am dong with my system and that part of the relationship is working.
-
A stick and slip where things are supposed to be smooth is not good,, I need to fix that.
I also ordered some new drive motors,, more torque :)
So far the system is doing what I would like it to do but the motors I have are not up to the challenge as well as the stick and slip issue already burning up one motor, hence the new stronger ones.
It is somewhat of a pulse system, that is half the time it is supplying a torque and the other half it is resetting with a greatly reduced counter torque,, so asymmetrical if you will.
When I have the motors and the stick and slip issue fixed I will then be able to run a better test and if I redo the reset system I can improve things that way as well, it should be an almost no cost thing to reset the system.
-
While I am waiting on the motors I thought of a way to reduce the load on my existing motor, it is only providing its input force for a partial motion so I came up with a way to use the rest of its motion to store some force and apply it when it is needed.
Running it this way the system appears to function, granted, as more of a pulse torque than a constant torque but the new motors will help with that as well.
There is lots of room for improvement but it appears that the basic concept functions in providing an asymmetrical torque from the system against the shaft it is mounted on between drive and reset.
-
Got and installed the new drive motor, those runs were very frustrating and come to find out the motor I installed was defective,, I did not think about it since these are new to me motors so they might be a little different than the other ones even tho they look the same.
Replaced that motor with another one, tests are still frustrating. When I load the system up one of my parts starts to fail, when I leave it lite on the load the one part works most of the time but not consistently and when I run it lite the mass of all of that part of the system accelerates fast enough to cause some system rotation,, so not desirable to run fast.
A part that was not designed for the load and friction value that it is running with this way is having some issues as well so I am going to have to make a part to deal with that,, the place I made something to take on that frictional piece is not doing its job,, the part has decided to move into its own area even tho it should be bearing down on the friction surface I provided.
With all the little run times I have had with the new motor I have had a few moments when all things line up and work like they should and I appear to get a larger torque one way than I do for the return in the other direction.
-
working the kinks out.
I have a few issues that I am working on,, one of them was rather stupid, I was over-driving the system and running into some of the limiters,, silly error.
When operated by hand the system provides one amount of torque one way and another amount upon the reset, the driven motion is the larger and the reset the less.
One of the things that is interesting is the interaction between the direction of force, whether from the input or the reset.
-
working on new ideas and am playing with one that is sort of like a circular Storks bill,, it is fun to play with the testbeds and might actually make a good design study for those who are in the mechanical world,,, fun things with force vectors and stuff.
-
I have been playing with a device that allows me to interact with a rotating force vector between two interactive components,, nothing big really but it is kind of fun to play with.
A few years ago I tried to come up with a Non-Rotating Lever, NRL, and as one might guess it failed. I also made a few other test-beds around some of the same concepts and well Palm showed a few interesting things but upon a sim being made showing the same kind of thing and then that sim being found to be flawed I just assumed it was some kind of unseen preload in my test-bed and left it at that.
I am slowly making the parts to attempt to make an NRL and if it does work then my other test-bed might of been flawed but not enough in the right way :)
-
My test-bed is really unstable so nothing I see from it is anything concrete,, it is trying to show me how I need to change it so that things only move like they are supposed to instead of with all the slop.
One of my assumptions is that if I apply my input force in one are I will get a different output force as compared to another area, that part seems to be present, funny thing was I thought it was backwards until I looked at and realized I was looking at my test-bed backwards,, that is as for as the force vectors go,, silly on my part.
-
this is an idealized pictograph of the relationships as I see them.
At first I thought that it would not be able to be realized in the real world, but with the appropriate acceptance of compromise I think now that it can.
Now to choose the compromise and design and build.
-
it is not as easy as I had hoped it would be,, but I am slowly moving forward.
I have made it as small as I could but still one part is bigger than my printer so I had to cut it into pieces and print them that way.
I will not be able to build a "wheel" with the way I am going about it for this test-bed, but I should be able to have an NRL and or a self lifting weight :)
When I have all the parts printed I will know if I am correct or still missing it.
The mysterious force of opposition :)
-
I made some parts that are close but I needed those parts to decide on what the correct parts should be,, confusing but it is how I work :)
I did assemble those parts even tho they are not super close to what I need them to be, and that was very interesting to play with since I was expecting to be wrong with my predictions on the behavior of the systems involved.
So far it works almost exactly like I predicted it should work,, now to design and print the real parts and then I can run some force tests with something other than Mr. Hand.
-
I have tried to incorporate a few "enhancements" into the test-bed I have right now. I have tried to include a system that helps control the mysterious force of opposition so that the output of the main system is even greater.
That system, the "enhancement" is not doing what I hoped it would, so far anyway.
The main system is showing and doing what I thought it would, that is it has a range of interaction where if using gravity the system would self lift that mass and pass through a transition point and then drop it back down.
I need to incorporate a control system for the transition point so that the system does not loose its relationships, the parts will just "pop" the wrong way if allowed and fall out of alignment.
I have sheared a few pins while testing things so far and so I need to keep the forces fairly small right now, but I hope if things keep going as well as they are and then I will change those pins and things so that a significant amount of force can be passed through the system.
-
I am a little frustrated.
I am trying to make changes so that in the end the system will be a constant exchange of an external force instead of this short range exchange. It is not going real well and I am running into the same issue with just ending up with a short range.
I have been blowing an awful lot of brain farts these past 3-4 weeks and to me that means that I am not appreciating what is actually happening within that short range, I am missing what the real interaction is.
-
Things are moving slow but they are still moving :)
Some of the print times are getting up to more than a day long,, it is a little scary at times how involved some of my prints have gotten.
-
eyes can be so misleading,,,
I have been working on making a shape, a particular shape and every time I thought I had it designed in FreeCad I looked at the thing and it just looked wrong,, so I tried again in a different way and so on and so on.
I have had this kind of thing before and I finally remembered that what "I" see as not level just actually be level from the parts perspective,, just as what I see as non round might be a perfect circle :)
So I went ahead and printed the part and it fits and does what it is supposed to do even tho to look at it you would not think so.
I have my first basic unit put together so that I can actually see and play and feel how things interact and so far it looks like it might do exactly what I think it can.
I will play with this for a while and then get into designing the parts more accurately and then see what it can do from there.
-
I think I just figured out a way to further multiply the input work.
-
Been busy :)
So things I am making are getting a little complex and yet very simple it is more in the how can "i" make it that is getting complex.
Now to try and remember how I used a few macros and programs to design a certain kind of part that as it turns out is something I needed to know how to do to make a part for this build,,,
Life is kind of funny sometimes,, the past time wasters might end up being what was needed all along for tomorrows next best thing,,,
-
I revisited my workup sketch and refined it to the point where I can use it to build a part that is needed.
Here is a pic of the file
Going this route I do not need to fiddle with the macro I was going to use.
-
some days I am slow at picking up what things are showing me :)
My present test-bed has a deviation from constant to provide a form of constant and I have been making my parts with that built in,, well if I don't build it that way then I am left with a 3.14 degree of change difference. My input travel and conversion travel have a constant difference, a small one granted but a difference non-the-less.
The 3.14 is just a coincidence by the way,, purely a geometric driven relationship.
-
things are interesting.
What I am working on I think I have all the forces accounted for it is just that they are not all equally distributed.
I am working through what I think should work and trying to maximize that inequality with the scale that I can build at,, make it much bigger and it would make more but I am limited on size so I am hoping that it is enough to overcome frictional losses,, well if it even works :)
-
With my current test-bed I can go in several directions for the rest of the build,, I needed to decide and commit to going in one direction so I decided to go with the less than route but a simpler build. Less than meaning not as much output as some of the other ways of doing things but still hopefully enough to run itself.
-
I have been building this test-bed so as to NOT use gravity,, that, I thought, would only confuse things, but I think for the ease and simplicity of the build I will use gravity in the output section.
-
I was playing with Freecad to determine the sizes of things I need to make and found something interesting as far as the ratios go,, the angle value can work in more than one way.
This means I might not need to build a part so tall.
-
I designed a part wrong, unfortunately I did not know I did so and then I spent 2 days working on changing parts to work with the error,, silly me and a waste of time and filament.
-
I have been going over the bad part design and even tho it does not matter if I understand why I need to design the part to be held in a place that is not the same as I think it should be held,, I keep trying to figure it out.
When I put it where I think it makes sense there is a lot of variation in motion, when I put it where it wants to go there is almost no variation in motion.
-
The idea of being able to make all my parts perfect is great but I am not going to be able to do that,,,,
What I am doing is making "room" for slop, as in if this is this sloppy where and how can I make it not matter so much.
I am also redesigning things as I go to have clearance,, a few parts collide and that is a problem of course :)
-
I kind of strayed from my path a little bit,, just a short sidetrack :)
Also I am recycling a few years worth of test-beds,, it takes a while to strip them down and stuff, that and it is also kind of fun to see just how intricate some of my previous builds were.
-
I was running with 5 sub-systems and decided that those were a little overkill and was going to change that, then I thought that maybe I could just make it all easier if I could only take 2 wheels of different diameter and have them rotate at a 1:1 with nothing else in-between them,, so no extra gears or levers or anything just the 2 gears geared together at a 1:1.
It took little bit of time but I have come to a way of designing this gear-set and am printing it to test. The difference is not that much it is like a 100mm radius to a 147mm radius. If that indeed functions as designed then I will re-design another part to make it much simpler and stronger.
-
I made a very crude part to test my thoughts and it work well enough so that now I will spend the time to plot out the profile.
-
It took me like 4 15 hour prints to get a crude shape that was close enough to let me play with the interactions I think are within this test-bed.
I plotted out what I thought should be the profile but that did not actually work so well, then I winged it and that was closer,, then I refined it a little more, again I am not making this perfect, I can't until I actually know what profile I need to follow.
As I continue on with this little adventure and put on some of the other parts, crude parts but able to show me the interactions, things keep looking positive :)
-
What I think I have done is to break the internal work transfer path.
If you have a 5:1 lever then on one side you have 1 unit of force traveling 5 units of distance, then on the other side you have 5 units of force traveling 1 unit of distance.
If you were to view the transfer of work then from the internal structure you would see both sides at the same time, the internal path of transfer sees the sum total of work transferred so you have (5*1)+(1*5) for total work internally.
Breaking this path then might allow me to add my own values internally, raising the system wide potential and then externalizing the resultant as a constant value needed to "balance" the internal change.
-
If then I want to put in 1 unit of force for 5 units of travel BUT I only want to take out 1 unit of force for 1 unit of travel then internally I would need to consume 4 units of force from one side and 4 units of travel from the other, constantly.
Externally I would need to consume 1 unit of force for 4 units of travel.
The unobtainium then is in figuring out how to break the internal path of work transfer.
-
Kind of interesting.
I changed the design of one part due to it not lining up with another part in the program, I printed this part and put it in the test-bed and it did not fit, the "adjustment" I made so that it looked like it fit in the program ended up not being needed, the part got real sloppy.
I took a leap of whatever and decided that the design may not "look" like it fits in the drawing program but that I will run with what I think the part should be. This is a long print time and some filament to use up but that is not that much of a risk.
In the program I placed the first part on top of the "adjusted" part and it looked like what I have in the test-bed, the first part filled in the gaps that the adjusted part created , the interaction profile in the program lined the two parts up with the adjustment, where-as in the program the parts overlapped each other with the first part.
-
A test part finished and I was going to wait until tomorrow but I couldn't resist and put it in.
This part, when I view it in Freecad, has a lot of the parts overlapping by a large amount and yet when it is in the test-bed and adjusted there is no such condition,,
Very interesting!
-
This morning I cleaned up the part and also double checked that in the program there are 10 parts that overlap, parts that should need to run through each other,, a physical barrier to operation.
I spent a little more effort in setting the part and there it is moving very nicely and smoothly and no overlap, no collision of parts
So here I am with a 0.75 : 1 distance ratio at a 1 : 1 turn ratio,, not saying that there is a gain there, I don't care if there is or not, the part is doing something else for the system.
-
I printed the parts rather thin, 3mm, and with the height there is a little warping that can happen with the shape that I am using.
With that said however, the "tweak" or adjustment to the profile of the part is very small and only a case of knocking down the corners a little bit.
What this part does for me is turn what would be the normal external work function and inverts it, that is it is now the internal work transfer component allowing me to access what is normally the internal work transfer component as an external function.
-
Well things are kind of nifty and yet frustrating at the same time.
I am trying to determine how much printer error and how much design error I am running into.
If I take the part I printed and install it a little "warped" after I clean up the contact areas things run fairly smooth and free,, but I have to force that "warp" into the part which changes one of the angles of interaction,, is that my printer or is that my design???? I think it might be printer mainly, not all of the interactions are maintaining the same distances when they should all be identical.
-
I found out why my part worked,, well it was a bunch of errors but I think I have fixed those and am printing off another part that should actually work, it does in the program and I hope in the test-bed
-
new part hot off the printer and thrown into the test-bed and that part functions just like it is supposed to.
-
It seems as tho I got a little over enthusiastic on the number of "teeth" I used on the gear-set, I used as many as I could fit in the space I have, well when I did that I also removed any slop room I might of had.
I need to re-do that and loose a tooth or two.
-
my method of working is slow. I am usually making slightly different parts as I go along,, I start with a normal style of part design and make a test-bed trying to look at what I think I want to see, then I change parts more to what I think it might do and then change them again until I think I get to what I really wanted to try.
With this test-bed I am moving even slower than normal,, it kind of gets annoying how slow I am some times.
-
Things are starting to look rather interesting finally.
This Work Transfer Function just might have some merit after all.
-
what I have right now is kind of promising.
I have 2 shafts that come into my system, one comes in from one side and the other from the other side.
To limit frictional losses I have used bearings every where I can and have no slipper clutches or friction items as such. I have tried to make the whole system move as easy and smooth as possible.
The shafts rotate at a 1:1 ratio and in the same direction.
If I use one shaft to rotate the system, and therefore the other shaft in the same direction of rotation, the system requires a fair amount of force. I am using Mr. Hand so the amount of force is unknown. When I use the other shaft to rotate the system the force required to do so is noticeably less.
I am still working on re-designing my parts to make this difference more noticeable and more controllable.
-
I am getting myself rather all mixed up and on top of that my base keeps wanting to spin but it supposed to be fixed in place. I was kind of lazy and decided to not "fix it in place" so I could just spin the whole system instead of getting up and out of my chair to look at things from all angles,, time to just lock it to the vise on my desk.
I can choose the interactions and change the behavior of those interactions by adjusting them,, that part is what I thought should happen so now to get things adjusted and see what works and what does not.
-
Things are getting frustrating, annoying, exciting, interesting,,
I have made for a fair amount of adjustability and that creates a few issues and one part that I did not make adjustable, per say, needs to be made adjustable.
The part that is not so adjustable is held by rubber bands, my thought was that I could see if the forces I think should be there are if that part shifted against the pull of the rubber bands, that does happen.
I have 2 parts that are supposed to have free movement in one direction but due to poor craftsmanship they are a little sticky and that creates issues. I set things in place with what I think the alignments should be, or close to it, and nothing happens easily, I tap on the table or test-bed and the vibrations allow the sticky parts to move as I think they should and things change as I think they should change.
This leads me to believe that the force interactions and motions I am looking for might be present within the test-bed.
I think this is all looking very promising :)
Now to design the next parts
-
This is a little perplexing.
What is supposed to happen is that I have two sources of force that are the same, the interaction via the device allows for the reactionary forces to be dissimilar and so to balance the reactionary condition I should be adding more force to one side. I am using mass and gravity for the source.
The thing is I am adding a lot more mass to the one side than I think I should and even still the light side is advancing the system and lifting the heavy side.
The drop to lift is also not as I thought which might be the reason for the increase in mass needed to balance, my thoughts were that it should be 1:1.4 but the distance change is more like almost 4:1.
I have not designed the parts to allow for to much change from a setup relationship.
-
I did not like the actual physical contact that two parts were having so I have been re-designing that contact, well I decided I wanted it to mirror another part interaction so I finally got a configuration that I think will do that for me, I am printing off the rough parts now to check on the setup.
-
Just going back over older setups and decided to actually run a force measurement,, a crude one but it is not what I thought it would be.
I have 2 parts that are not the same but they move at a 1:1 rotation value.
One connection point for my scale was at 90mm from its axle and I applied 2N of force.
The other connection point for my other scale was at 145mm from its axle and I matched the pull of the other by applying 1.8N of force.
I used other force values,, 3 to 2,, 6 to 8.4 and through some parts of the exchange it is almost a 1 to 1
My interaction is not a constant thing so there is variation of the exchange distance between parts which is why I chose distances from the associated axle instead of the ends of the parts, trying for an actual leverage or torque value as seen by the axles.
-
I did not design one of the parts to move, it is actually sitting on a table and the other part pushes down on it while things are in motion so that part has more frictional losses than it needs to.
That part is the one with the 90mm measurement point.
I am a little annoyed that the system looks like it might work this way, I did not think it could and went the other way with my build.
By the way the larger force numbers go with the 90mm point and the smaller ones with the 145mm point and both points rotate.
-
so I pulled the test-bed apart and put it back together so the one part was on bearings.
I changed the measuring points and got them closer in length.
I ran the pull tests using one scale on one side and the other on the other then I switched the scales and did them again.
135mm to 155mm
300 320
600 680
1.2 1.45
200 240
600 680
1.2 1.48
-
I was playing with the variance by position and I almost got it down to 2000 to 2000, so very close to a 1:1 but then I also had it at close to 1600 to 2200, again that is on 135mm to 155mm
Now to think about maybe redesigning this setup to work this way. I am not sure if the small gain I have is worth the effort,, it looks like it might average out to something like 120% to 130%, then again when going full rotary it might end up at 0 gain.
-
Just playing some more.
I think I have another method for interacting with the base unit that may even be better than the 125% way of interacting with it.
-
Very frustrated,,,,
I have to figure out again how to make a part so that I can print it, it was a convoluted process of making this this way and then taking that and making it something else and then being able to make the part so that I can print it,,,,
In the meantime I remeasured my pull points and this time set them almost identical to between approx. 150mm to 155mm for each point.
These points are showing approx. 133% out compared to the in.
Anyway,, once I can make my part again then I am going to downsize the build a little bit and design it for this actual usage and try and remove most of the slop.
-
I just did some basic math to see if what I have measured is reasonable.
The range of relationship I should have goes between 1.26:1 and 1.41:1 so an average of 1.33:1 which is close to the range of values I am measuring using simple spring pull force scales.
-
I am redoing the base system to be used this way and I am increasing the multiplication up to almost 2:1.
I will have a small issue with the structure of the system being rigid enough to withstand some of the force values,, in other words I might break it fairly quickly.
I am looking at several days or more of print time and I am hesitating because after so much print time if I break it before I can get some usable data maybe I should try and make it more robust, or maybe it will be just fine,,, or ????/
-
A small note,
I am using my existing test-bed to look into these forces. I have added a part to that test-bed to help hold things together but I can not keep them held so well and I have been moving my lock pieces even tho I am using 2 lock screws per lock part as well as my parts start to deflect, or bend.
I am thinking that I might be able to use just enough force so that the frictional losses are not a very large part of the transfer work part.
-
some days I think I sit and spin because part of me knows a better solution.
After thinking about my setup I came up with a simple way to remove these pesky force issues,, a very simple solution. This solution does not require too much re-design and only a few more parts per system.
Leverage and torque are ratios of force and distance, so 1Nm is 1 newton of force at a distance of 1 meter, this relationship then has 2 extremes that do not play nice, those are, an infinite force with no distance or an infinite distance with no force. I am going to use these conditions to my advantage :)
-
Designing my parts one way,,
I need three of the same parts in this setup and one way I can go with those parts will take me 65hrs of print time per part!
I think I will go another way :)
-
changing that design to go another way the part will take 30 hrs to print,, still need 3 of those, and then there is the timing parts,,, this is going to take a long time to print either way I go about it.
-
I think I am going to make a test-bed that does not go into a full continuous rotation but rather a reciprocating lever function.
This will make the parts smaller and easier to print, probably small enough for almost any printer to print them.
I could then take the output and run it through a mechanical rectifier if I feel the need to have a continuous input oscillation converted into a continuous output rotation.
-
Well,, that was frustrating. I started the design and slowly got more and more carried away with until it was another huge elaborate part that would take 33hrs to print.
Next try :)
-
I am trying to design my test-bed so that it can be printed on almost any printer.
In going through this exercise I think I have figured out why my other test-bed was providing me with some funny values that I did not like.
In my system I can pass the work through the the system in several ways but basically I have two sides to the system. I can pass the force value at a 1:1 or I can pass the distance value at a 1:1.
If I were to use one side as the input and pass the force value at 1:1 then the input side will pass 1N of force by a distance of 1m, the output side will provide for 1N of force by a distance of 2m, if then I were to use the output side to reset the system I would need to provide 1N of force by a distance of 2m to recover 1N of force by a distance of 1m.
If I were to use one side as the input and pass the distance at a 1:1 value then the input side will pass 1N of force by a distance of 1m, the output side will provide for 2N of force by a distance of 1m, if I then were to use the output side to reset the system I would need to provide 2N of force by a distance of 1m to recover 1N of force by 1m.
What I am designing right now is a partial system and so these exchanges can be easily observed but a full system provides for a unique situation where there is no reset required, the end of one exchange brings the next exchange into its starting position and since there is no work transfer being performed from the next system while it is not being used, only moved, there is no cost to moving it.
This is my concept anyway, so far it looks functional but I will not know for sure until I finish the test-bed.
-
So I decided on the fishbowl part and am now designing the table,, I am printing the fishbowl :)
-
Wellllll,,,
The table leg is going to take 84+ hours to print and it is 196mm x 252mm x 150mm
I say + because my printers do not always end up taking the same time as the slicer says it should take,, the printer usually takes longer
-
as the print goes on, and on,,,
If this does provide what I think it should then I was thinking that I could then actually cut the parts in half, only using one half would make the print size small enough for most printers that can go at least 150mm high.
A complete system is another matter all together, it has a few more parts, like 6 full subsystems,, I am only making a partial subsystem for this test-bed.
-
Still printing,,,,......
I have designed the other 2 parts I needed to and when my other printer is done with its print job I will print those so then all I have to do is assemble :)
-
still printing,,,,,........
-
This is really interesting.
My system breaks the pathway of the work transfer function, sort of.
I was looking at the rendering of my system that I am printing and noticed something that I thought was kind of there but did not overly think about it, so I took my current test-bed and set it up so that I only had a single interaction just to see if what I saw in the rendering was actually there.
The interaction of the input to the output has a force a path and a direction, several actually, and it is showing that a pull is not the same as a push, it is showing that if I wanted to have a longer distance of interaction that I would need to transition a pull into a push at the peak gain in leverage for the input over the output.
So, the system is showing in the real world that when the input is pulling the output through the beginning ramp up of leverage advantage it *appears* as tho the input is moving the output a further distance, it does actually in a sense, and then if I keep the input pulling on the output the input then is traveling a further distance than the output,, IF however at the peak I change the input force from a pull to a push while maintaining the same direction of interaction then the input is moving the output a further distance still.
Even tho I can see these changes the changes themselves are almost inconsequential to the work transfer from the input to the output, what I mean is, if I focus on watching those changes I can see them but if I only look at the rate of rotation between the input and output that stays as a constant 1:1.
This is getting to be really interesting.
I will have to wait until I have the test-bed finished to be sure but I think I will need to print another "adapter" part so I can change the input from a pull into a push but I should be able to measure a difference in distance between the two sides of the leverage function, that is as the leverage is growing up to a 2:1 compared to when it is reducing from a 2:1.
-
Still printing,,,,....
-
4 days 5 hrs and 42 minutes,, longest nonstop print I have done so far.
-
basic clean-up and parts fit, motions and all that so far are looking very good. The input rotates 45 degrees for this test-bed and the output rotates 45 degrees as well. When I try and continue the input as a pull past the point of change it rotates another 45 degrees and the output rotates maybe another 15 degrees, this behavior is what was predicted.
Even without all of the supports and lock pieces in place this test-bed is way more stable than the other one.
It looks like I am off on one of the contact surface faces, it should be touching but I have a gap of about 1mm or so, I may decide to print another face that closes that up,, I will see how it goes.
-
I am rather confused by all this.
The test-bed works fairly well as far as being smooth and all that so I think frictional losses and what-not are reasonable.
I designed this one to provide for up to a 2:1 and it is only a single tooth and only works for a partial movement.
When I keep the system still and pull on both of the spring force scales they will both most often read the same force, I can then allow the system to move and maintain that same force reading OR I can take the output side and increase the force while the system is in motion, if then I bring the system to a stop there can be a difference of force or it most often will drop back to the same values again.
I have not noticed the input side going higher while doing this, but I have not tried that way so much because the output side seems to make the change. I need to work on that part more.
I have not yet actually checked the distance of motion but since both the input and output pull strings are using the same diameter pulley and they both are rotating approx. 45 degrees I am just assuming that they are moving the same distance.
When I move past the change position there is a large difference in distance of change and a large difference in force readings.
When I hook the 2 force scales together by themselves I can not get a difference in force readings whether I am moving them or holding them still.
So far I have been using about 600g of pull force on the input while allowing things to move and the output side will read a force of opposition of 600g and up to about 1.2kg.
This test-bed has not answered the question of how and why yet,, I was hoping that answer would just sort jump right out at me.
-
I am running with the assumption that I have some kind of bind situation occurring while I am stressing the system.
I can go from the start position and pull on both spring scales the same amount and make a motion maintaining a close value between both scales. I can also choose which one I will pull harder on and watch its scale go up but see very little increase on the other scale, within a reasonable amount of increase, like 150g.
I "broke" one pathway with this build, I thought I might break the test-bed if I did not, I have since printed off a part that puts that pathway back into play and things appear to have changed somewhat in behavior.
Now I am printing off some parts to close up some slop.
-
Not a bind,,
I made an error in design and the system is setup so loose that things kind of move the wrong way.
I made a "fix" for most of that and now I am getting the force differences but they are not as much as I thought they would be, so at its peak difference right now it is only 1.5 :1,, I also need to check on the string pull distance to make sure that that is not also 1.5 :1 which means I will need to build a rig of some sort to check that.
-
I just had a funny thought,, what if I took out the path locking part and used my open path part again,, well I did and the pull force went back to being the same.
Interesting. Some minor differences could be had but it is back to depending on which spring scale I pulled on harder first.
-
So I was wrong on the closed and open path, still figuring that one out. The 4 day print job was somewhat not needed but it did show that I either made that part wrong or that removing that pathway is problematic for my desired end goal.
I printed another part so I could actually check the push side and it works the same as the pull side. I made that with some slop but unfortunately that part would require zero clearance.
I am getting a max, at this point, fairly reliably of 1:1.5, that is my input is going from a 1:1 and ramps up to a 1:1.5 and then it goes back down to a 1:1. This is happening while the input and the output are rotating in sync at a 1:1.
Now to ponder if I want to build a full system and close up the slop as tight as I can,, or if what I have is enough to satisfy my curiosity.
-
So I thought about stuff, and well I got sidetracked :)
I took my locked path part and tried to lock the path with the zero clearance I need and that worked nice a few times but as I was using it that way and then trying to look at the scale readings the force transfer was getting to be less and less of a gain, that had me puzzled until I noticed that my lock piece was actually moving and allowing the clearance to grow larger and larger, move things back to a "tight" fit and run it back and forth a few times and the lock piece moved. It takes a fair amount of effort for me to move that lock piece and yet the system is moving it which means of course that I must then lock that lock piece,, or maybe don't bother and do something else.
The first time I was able to watch the scale readings they did get up closer to like 1:1.8 I was seeing around 650g on the input and around 1200g on the output peak difference which is better than the 800g to 1200g I was seeing before adjusting the lock piece.
-
I took the lock piece and screwed it onto the lock hub so it does not move, then I used a screw to set the distance to get rid of the clearance but I am only stopping that in one direction, so there is a push against the set screw but it does not pull on it.
What is interesting is that this changed things, so I can pull harder on the output while moving the input or I can pull harder on the input while moving the output and this little subtlety is a little problematic, if I do not allow the same motion between the input and output spring scale the readings appear to change which makes no sense, worse is when I use the push side I can pull on the input scale a huge amount with little to no change on the output scale.
The only other thing I changed was to add more tension on one of the shaft bearing sets, to reduce the shaft deflection or run-out.
I need to get better at moving the spring scales and then I might be able to figure out how, when the system is in the same basic position, I can get very different readings that seem to be dependent on how the scales are moving.
If the system is as it should be, that is just a 1:1 transfer since the distance is the same then the force should also be the same,, then how am I getting readings that do not show that for every reading?? How can I show say 1200g on the output and then move the input up to 2000g with no change in the output scale,, I actually had that down to 800g on the output and over 2000 on the input.
The thing is I can not feel anything that is sticking, even under load, and yet I can bounce the force on these spring scales a lot without that reflecting on the other scale,, makes no sense really.
-
I built a string guide plate so that Mr. Hand has little to do with things and I turned all forces into vertical forces.
I placed an approx 500g mass on the input string to get rid of that spring scale.
I used the spring scale on the output side and tested for different spots.
For just about all static conditions the scale read approx. 500g, the scale weighs 50g.
While moving the scale could read more but I would need to keep accelerating the scale,, f=ma after all.
The arm lengths are a little interesting, the output is fixed at 125mm the input changes from 85mm down to 65mm.
Both strings are on the same 55mm radius for the pulleys.
Both strings move the same distance as each other, in sync but reverse directions.
I think I have everything adjusted fairly close to where it should be by the design,, as it is now.
At this time and this setup then this system is doing what it should and only giving a 1:1 transfer of work.
I tried playing with the push side,, things need to be adjusted for that side before I can play much with it but the little range I did have did not seem to be that much different than the pull side, things just popped much faster past the end of range.
-
In a nutshell,
I have observed that when Mr. Hand is supplying the input force that the systems I have built CAN have a different work in than out, the question is how.
When using Mr. Hand the force is only constrained to a tension in the string which is wrapped around a pulley as such that the tension creates a tangential force which is converted into a rotary motion.
I built a pull string guide system so that the forces on\from the strings were fully constrained and in that condition the input work equaled the output work, as is predicted by standard practices.
The guide system is attached to the base system, as such the system is then closed upon itself.
The task now is to find out which of those closed constraints I need to open and in what fashion so that the system reliably shows a change in the work performed between the input and output.
-
I think I am identifying the causality,, not so sure on how the system translates the forces and motions and stuff but the area I am looking into does seem to create a secondary feedback that can either add, subtract or stay neutral depending on the angles of interaction and directions of forces.
I have one possible condition that is interesting. This condition is where with one direction of change of force creates a condition where the system seems to want to find balance after it has lifted a weight, reverse the direction of change and for the same change for the system the weight would be dropped twice the distance,, so I am thinking that if I designed that stage of operation so that the lift condition is converted into a force of rotation as such that the change in potential rotates the system to maintain the relative height of the mass instead of lifting it that this might provide for a reset condition where I would need to disconnect the mass from the pull string, move the system back to the start position and then reconnect the mass to the pull string and allow the system to move, with force, back to the end position.
-
I printed off a few parts to start testing using the test-bed this way.
One thing I have noticed is that the system responds differently when using what I have labeled the pull side and the push side.
It looks like, not tested yet, like my input force and distance are not the same as the output force and distance.
I will need to print off a few more parts to make testing more straight forward.
-
These are just some crude measurements of force and distance
I have an average force of 77.5g moving a straight line distance of 95mm on one side
The average is between the relax to move backwards being 60g and the increase to 95g to start the pull forward
I have a force of 185g moving a straight line distance of 28mm on the other side
-
I printed off another part so that the 2 distance values I am looking at for this setup are the same.
With that, one force moving the same distance of change as the other force,, those are 185g and 200g.
With consideration for friction and stuff I would say that they are very close if not close enough to call them a 1:1
I have another "out" method I can use and when I use that method my input force is 185g but my distance of change is still needing to have some parts created for,, I can see the distance Mr. Hand moves but I need to measure it correctly, that distance is less than the output distance of change and when I have those parts I can then measure the output force and distance of change.
-
That does not make much sense.
I am walking down a dark hallway of the unknown and I am trying to place lampposts along my path and using the light to try and see what the next step should be,, or could be.
Mr. Hand with the present test-bed setup is supplying a few things, 2 forces and 1 distance.
I can replace the forces with a mechanical system for each one that would not cost me anything leaving Mr. Hand only supplying a change in distance,, and a catch and release mechanism.
Think of it sort of like this,, a pulley with a string over it and equal masses on each end of the string, Mr. Hand is holding up the pulley, this can easily be replaced. When the system is active there is another force 90 degrees to Mr. Hand that does not move that Mr. Hand is resisting, this to can be easily replaced.
Now Mr. Hand needs to move one of the masses up and down and since there is an equal mass on the other side of the string Mr. Hand is only supplying the change in distance.
My end goal is to hopefully make an NRL, Non Rotating Lever.
I have a system that can supply an infinite distance of change, I am building a system that should be able to supply an infinite force. These 2 components are the WTF that is customarily trapped as an internal exchange of potentials but with the systems I am trying to put together this internal exchange is being allowed outside so if I place a potential of 1N and a change distance potential of 1m then the system should respond and externalize that exchange as work performed from the supplied potentials.
-
I have been using what I have and testing.
Right now I have the approx. 185g mass dropping approx. 65mm. When I raise the mass and set the system so that that is the start point and then release the mass the system will rotate CW, the system then, with how it is setup, will create an approx force equal to 300g moving approx. 38mm.
This is close enough to say that this is a 1:1, the work I put in to raise the mass is equal to the work I can recover by letting it drop the same distance.
When I break the internal path and externalize the WTF the system responds by rotating CCW with some force value, not measured yet, and the mass drops some value, not measured yet.
If the system were designed so that there would be no prop in the mass then takes care of that reset part.
So then what it looks like I can do is put in the work to lift the mass by itself with the system in the end position, then externalize the WTF and have the system move with force into the start position, internalize the WTF and release the mass and let it drop 65mm moving the system with force into the end position and recover that work into something so that I can then lift the mass.
Just for more confusion, I need to lift the mass and then move it into the horizontal position where the system would be when I am going to release the mass and let it drop, a horizontal movement has no work needed for the mass.
I will be making more little changes and testing trying to get the externalized distance so that the mass does not drop.
-
I did a quicky,, and I am able to stop the drop by design.
The interesting part is where this force is favorable and where it is not.
I have a partial range of motion where without the mass dropping the system moves with force one way, I go past that point further and the system moves back the other way with force to the balanced area,, apparently.
The drop condition is still the same but now it drops further.
-
What I am finding is that these setups are trending into a 1:1 so with the one funny setup I am expecting that there is going to be some change in height that I am not able to measure yet and that that will account for the perceived work out.
Using all the same parts I can get different force relationships, right now I can use the same 185g to create a counter force needed of 600g,, just by changing a path
-
I think I have identified one of the things Mr. Hand is bringing to the party.
Take a pulley and tie a string to it, wrap that string around the pulley a few times CCW, pull on the string. The first thing you should imagine is that the pulley will rotate CCW, well of the pulley does not rotate then the only way the string will unwind is if the relative tangential point rotates CW,, this is what I think Mr. Hand is doing that is adding an influence to the test-bed.
The pull point creates a relative point in space away from the pulley as a tangent and the center of the pulley and that creates an angle that could be used to determine the amount of rotation force seen, or you could use the leverage ratio to find the same thing.
-
A simple view I keep in mind is a 4 quadrant conversion thing.
The quadrants are for distance and force, so in the top left quadrant you have a +,+ in the top right you have +,- in the bottom right you have -,- and the bottom left you have -,+.
We work in the top right and bottom left, so an increase in distance comes with a decrease in force and a decrease in distance comes with an increase in force. These 2 quadrants work together to give us an equal and opposite interaction.
The work transfer function uses the other 2 quadrants, so on the one side you have an increase and an increase and the other a decrease and decrease. These 2 quadrants work together to give us an equal and opposite interaction and is an internal function, or a closed system that we do not to play with.
My one part allows me to at least "see" this internal function.
-
I think I have got something figured out that is very interesting.
I think I can give up the distance for increased force without an increased cost to me.
If I use a setup similar to my worm-drive system.
My worm-drive system allows me to use a larger worm gear than the driven gear as well as reduce the frictional contact losses and make it a bi-directional system, that is each gear can drive the other gear.
Some may argue that my usage this way is not at all the same as my worm-drive,, but I think of it as similar.
I will print a part to test this thought and if it works then I will have a limited rotation that will then be reversed, so an oscillation. If that works then it could be made to be fully rotational.
-
Another interesting thing that this system appears to be doing.
I have 2 gears that are geared together and they utilize a continuously rotating force vector for the exchange.
If I use them with the input to output the same the result is a 1:1 turn, torque and velocity BUT if I use them another way, same setup gears are locked together so all things appear the same except that the force, rotation and velocity are varying. One gear almost comes to a complete stop while the other gear is almost free to rotate, then that gear will accelerate the other gear then decelerate that gear and it will almost come to a complete stop,, rinse and repeat.
There are 2 points in the interaction where I can choose which way it is going to operate and the change I think I see is in the internal direction of force transfer, that is whether it is a push or a pull transfer while maintaining the same input to output direction.
-
actually there are 4 spots where I can choose.
-
This is going to be interesting.
I need to combine what I am doing with another thing I made, my worm drive.
They both do something in the mechanical exchange but in a sense they are opposite each other in the conversion process. They do not need to be 90 degrees to each other per system really, it is in how they convert motion and forces. They both work out to be 100% ideally efficient and close enough in the real world, so what one sees as a transfer function the other sees as a non-changing point of observation so it can make its own transfer function.
-
playing with a bunch of settings and stuff I had a thought that maybe what I was doing was not what I should be looking at, where and how the force should be applied.
I printed off a crude test part to place the input force where I think it should be now, instead of how and where I was trying to apply it.
My input force is so much that it is flexing all my parts, that coming from almost no input force from me due to leverage.
I can not stop the motions well enough to take any kind of measurement, I need to make a couple very rigid parts. I need to stop or severely limit the amount of flexing.
-
I think I have a crude conceptual understanding of what is happening.
The system I have is basically converting a reciprocating motion into a rotary motion, however the two main parts each think they are rotating. This sets up a continuously rotating force vector between both parts and this rotation is the conversion of force and distance, so as the force goes up the distance goes down and vice-versa. Each system is seeing their own conversion and the resultant is the throughput, which is in my case a 1:1 but can be any ratio desired.
All of this sets up, or creates, a third virtual conversion reference frame that mixes the two others together. This is a work transfer function and it is this part that I am now interacting with.
One of the things that is interesting with what I am playing with is my input force needs to be in the same direction whether the point of observation is moving up or down, using up and down as the direction of reference then my force is always down or always up even though the motion externally is changing between up and down. Doing it this way the assist is itself rotating, so I get a 90 degree rotation where it is adding to the throughput and then I get a 90 degree rotation where it goes from some assist to zero assist and then back to some assist and this happens as the relative up or down is changing direction. This means in simple terms that I would not want to have the assist in play during that part of rotation, I could but it would just be easier to turn it off through that portion and if I were using gravity it would almost turn itself off naturally.
The other thing that is interesting is that with all of the constraints built in I would not notice much except for a shift in the phase of the two main interactions, they could go a little out of sync with each other until the physical constraint is reached of the designed system. The assist is actually the force required to keep the systems in sync and held against one side of the physical constraint.
-
still working through these funny almost virtual relationships,,
-
This is a visualization of the path a point on a flywheel takes when the axis of rotation is rotated 90 degrees at a 1 to 1 ratio.
The one on the right is a single path, the one on the left is multiple simultaneous paths
This is the simplest visualization for the force vector rotation as all this is happening. You could then imagine that the if flywheel is rotating many times faster than the axis is being rotated that would "smear" the path taken in space fairly quickly so as not to be overly discernible.
If the angle of change is not at 90 then the path is shortened and the rotating force vectors are also changed.
What is interesting, and what I am trying to play with, is that for an angle that is not 90 you can have 2 different paths with the same total change covering the same range of change, so opposition may take a different path than causality with a different vector rotation value.
I seem to be having issues with uploading attachments right now.
-
here it is
-
here is one at 30
-
This is an alternate visualization for the exact same condition.
-
that last one was actually for a 2:1, I was looking at the smear.
-
I am going to share an extremely simplified observation.
If I evaluate a simple lever system and supply the input side with a force of 1N at a distance of 2m from the pivot and then I also supply a force of 1N at a distance of 1m from the pivot in the same direction as the first force I am sure that everyone will quickly see that the lever reacts with a value of 3Nm of torque.
-
Another simple observation.
Work performed does not care if it is "taken" out of the system and used or if it is "taken" out as losses, it is all the same.
-
annoying thing when you keep designing a few parts wrong, print them and try them and that is when you find out you goofed.
I finally took the time to draw the parts freecad and their actual relationships to each other to get them correct.
-
got those working correctly now to modify a few more parts so that they fit and function correctly.
Funny thing,, I sometimes do things that make no sense, or, for the most part, it would seem that doing it this way should be no different than that way, but sometimes things behave differently and then I have to figure out why. This is where I am at with this test-bed. I could make one part much easier but then it does not behave the same as the way I am using it within my test-bed even tho in one sense the end result is the same thing.
I realized that I need to make a similar change in method for another part, even tho the end result is the same, doing it "another" way should make it behave differently within my test-bed.
-
forgot to mention that after I got those parts correct is when I realized that that is not the way to do it.
Sometimes you want a torque and sometimes you want a leverage, using the wrong one may provide for the same outcome or it may add things you don't want or not add something you do want.
-
An observation.
I choose the relationships I want to run with and then design the ideal path for the rotating force vectors. This path does not have a dimension to it, it may have a length dependent on angle and displacement but the actual path has zero 3d value, it is a moving scalar.
When I add a dimension to that path things break, that is there becomes a point where the path needs to have a negative distance. In my actual test-beds I have blamed it on slop and printer error and stuff like that as well as issues with the program I use. The larger I make the system the smaller that point becomes and I have just removed material to clear the issue. This point appears as a binding point in my test-beds and I did not think about it at all. That was until I tried to design the system smaller and my program was giving me a strange shape that at first did not make any sense. I have had some issues with the program making the part I want but I figured out a way of dealing with that and just went on my merry way but not when I scaled things down except for the path dimension.
So when I take this ideal dimensionless path and add distance to it, around the point where the relationships change to an infinite force and zero distance with the oppositional component of zero force and infinite distance, this negative distance comes into existence. I think that the force side might also go into a negative value, not to be confused with direction of force, but I have not looked into that.
Anyway, in my test-bed I am simply going to omit the sections where this occurs and provide with enough overlapping points that the system stays balanced.
-
I am choosing my relationships and dimensions,, so just to be clear these setups and conditions are special use cases and no conclusions can be drawn except that for these conditions I have chosen I can have these interactions.
-
I am trying an old design for a part and while those parts are printing I have been playing with my test-bed as it is.
To date my tests are showing for the subsystems that they create a 1:1 relationship when I use them one way, a more standard usage. I have been playing with combining 2 of the subsystems and seem to have an interesting reaction when I try and run things backwards if you will. What I mean is, that using weights and I try an hold one up while having the other counter-balance the elevated weight I can get into a setup where my moving force is almost nothing but the needed torque to hold the elevated weight up is a fair amount, the setup wants to drop the elevated weight and to stop that I need to supply an input torque separate from the moving force and when I get the torque close enough and the moving force in the correct relationship, the moving force drops way down while the holding torque drops a little. My assumption would be that when the correct conditions are met that I would have a holding torque that needs to move along with the moving force and that moving force should drop to zero leaving me with a "sink" or, to get no work out and allow a motion I would need to supply input work, the holding torque and the rotation of the system just to maintain the height of the elevated weight and have it move horizontally.
-
To add a little more.
I am using a motion input and a hold torque, if for nothing else it identifies the 2 inputs.
As I change the relationships I go from high hold torque and a lot of motion force to a lower hold torque and an almost zero motion force, I keep changing the relationships and the hold torque goes down to almost zero and the motion force goes high.
My 2 inputs are making the input needed to hold the weight up and move it and they are splitting that needed work dependent on there relationship to each other and not the final work needed. The final work needed describes the total work that the 2 inputs need to provide.
-
I'm just going to put this out there.
A square does not collapse symmetrically as seen by the center.
Think about it, pulling the 2 cross corners together you take away 2 distance units for an increase in change of 0.82 distance units.
-
What I think I am seeing is that the very methods we use to constrain and design a system is what forces the system to behave the way it does. If we want the system to behave differently then we need to use a different methodology.
My test-bed, as it sits right now, shows me where the zero distance to infinite force is and why, if that is allowed to exist in all aspects, it stops the system from doing anything different than what would be expected.
My test-bed, as it sits, has demonstrated that if I use a standard method of interaction that I get the anticipated 1:1 transfer.
My test-bed breaks the force triangle in one way while still being a mechanical system with everything connected, this allows for a new "path" to exist. I think I understand another way to break the force triangle and allow for another "path" to exist through the very same parts.
Using an alternate method then, and not constraining all points as is usual, should allow my test-bed to behave in a non-standard way.
-
I am making parts so that I can feel what is happening and then make adjustments as I think I need to.
Here is the funny thing, I need to make my force input as close to only force potential as I can but I know that I can not make it perfect. What I have so far is an input force system that has far less unwanted work needed than the work I should be able to take out,, I think.
I am not using constraints to keep things lined up perfect, example: I have a bottom part that needs to be free to rotate on its axle, well I am using a large wheel that is free to roll on a hard surface with the axle having a bearing for the part to rotate on, so I am not keeping that point in perfect alignment but rather looking at the force that the large wheel is trying to shift with, and then making assumptions that to move that wheel is the work input I would need to provide.
-
Still making parts,,, but with what I have I tried to see if I can even find a positional relationship where an input force only transfers as a force with no motivational component other than to move a "lever".
Well It may not be ideally what I am looking for but I can pass a leverage into the system against a force of opposition, keeping the lever balanced by using my system to supply the other half of what the oppositional force needs. I then move that "system" in a rotation fashion and the output is what I think I am looking for,, the not so good part is that it is a fine balance point that can be upset and I am looking for a stable condition that is not easily upset.
-
Sometimes trying to follow all the twisted paths gets a little much, but the more I look at them this way the easier they get to follow.
One thing I have not figured out, and most likely I never will, is how I can push against a wall directly and have that push move me closer to the wall or ,conversely, pull on a rope attached to the wall and have that pull move me further away.
-
I have been working on a re-design of my force input section. I wanted to reduce the parts count and keep things as simple and uncomplicated as possible.
I had to give up on some of the possible force amplification to go this way but that should be no big deal.
With the newer force section design I should be able to incorporate my input force and my input motion into one input, one aspect making force and the other making rotation.
I have built 1\2 of the force section and tested it in a static kind of condition, that is, I raise the force potential within the system and after that is done there is no rotational force from that while holding the force potential. This might work fine with only 1\2 the force section but then I would need to provide 2 input sources instead of one.
-
That design worked for what I thought I needed but it turns out that was not what I really wanted to do, so it did not work for a gain.
I have noticed a few odd things however, one is that when I assemble it with one part or not that the system behaves in a similar manner but not exactly the same. I am meaning that I "can" get the same throughput either way, but in slightly different paths.
I made another part to test some force interactions and that part does something interesting as well when I use the part or not, again similar but not the same and if I use that test part with the other part and spin it up too fast the test part lifts up and off of the shaft, the faster I spin the test part the more resistance I need to add to the output shaft. I have a weight sitting on the edge of the test part and my input rotation is 1:1 with the output rotation.
-
My input is via a long "T" handle Allen wrench. When I start spinning the system I am putting in a lot of work and using one motion or path of input. When the system is spinning up at some point I shift from making a circular motion with my hand at an almost perpendicular relationship to the test part to a centered but angled motion.
The Allen wrench is not fixed to the test part but is pushing against one surface of the test-part so as to cause a rotation of that test part, which it is free to rotate on the shaft it is on.
When I have it spinning up at some value and the test-part starts to lift up the shaft and I am using the angled motion my input feels like it drops way off and as I try to spin faster I need to add a larger resistance to the output shaft to stop it from spinning faster which would lift the test-part up and off of the shaft. Done that a few times now.
The whole test-bed and mount table shake a lot since it is out of balance due to the weight on the test-part.
Since all this is being done with Mr. Hand It does not mean much of anything, but it is only an interesting observation.
-
I think I understand what Mr. Hand is doing.
Initially Mr. Hand is spinning up the weight, then CP is building and creating its own input force, this new force quickly outperforms the weight. After this point Mr. Hand is no longer primarily supplying a force of acceleration to the weight but rather keeping the CP and the system in sync.
-
I do not like using a velocity related force as a drive force, that is I do not want to use CP as the output force source because that means the faster I spin the system the more force the system puts out. I would rather have the output of the system dependent on the amount of force I put in so that it all can be controlled for the application that it is designed for independent of rate of rotation.
I have built test-beds in the past that I could use to replace Mr. Hand with the current test-part and test-bed, the basic need I understand. While I was going over design ideas and I was going back to an old test-bed I built, I realized that the issue I ran into with that test-bed is exactly what is happening within this current setup.
It all has to do with the difference between torque and leverage and the work transfer function that each uses.
A simple solution I came up with for that old test-bed, I used, but I did not like so much because it limited my access to some of the interactions I wanted to play with. Looking at the setup on my screen I see that there are other solutions, one of which I used but for a completely different thing, and one that is almost easier than the one I used.
I made this other test-bed that allowed me to take a non-circular motion and convert it into a circular motion, just not the way it is usually done.
I am going to combine these things from the past into a single device for the input section of my current test-bed and hopefully I will be able to control the input force independent of rotational velocity, but at the least it will allow for a synchronized input to output timing without interfering.
-
I opted for a chain and sprocket for part of this, I thought it would be the most economical way. Well when you make some silly mistakes it gets to be frustrating when what should be a nice smooth movement ends up being hard and messy.
When using a chain and sprocket it is important to chamfer the edges of the sprocket, not just make the sprocket 0.5mm thinner than the chain, then I was off by just a little bit on shaft spacing so I had to add a 1\2 link into the chain, and I am using plastic printed sprockets on printed shafts going through metal bearings and using a metal chain,, guess what bends a little bit,, guess what has a little run-out,, silly mistakes.
-
Getting frustrated with run-out and such.
I do not want to spend a lot of money buying sprockets and stuff so I have to come up with a way to print my parts and get them close.
My chain goes to tight in one section of motion and then too loose but when you are trying to keep things lined up that does not work very well.
-
I changed the way I was printing the sprockets and it looks nice, however I am going to be transferring a lot of force so I thought it would be a good idea to thread some steel rod through the shafts. That sounds like a good idea but the issue is that it allows the shafts to warp a little bit and then the run-out is huge,, so I guess I will need to just plan on less force for now.
-
I need to design in a point of failure that will be easy to replace. I need this due to the limited force that my plastic parts will be able to handle and I would not want to break a part in the middle of everything that would require me to disassemble the whole thing to replace.
I also need to be able to amplify my input work enough to more than compensate for my frictional losses and show a gain.
-
This is getting somewhat intriguing.
I have the system almost complete, the new part, and I made 2 crude alignment type of test parts to see about the angles and stuff. I should have a change in the force function depending upon the angles.
As I am playing with this there are a few positions I can set things into that behave a little different, that is in one setup the reaction to the"input" force drops and the system will make a fast change so I am thinking that at that relationship then if I were to keep shifting the reactionary "input" force, which has dropped, the system would keep shifting. This is close to what I am trying to do.
Another one would be using CP where the CP hold force is converted into another output work function.
-
I should also mention that I made these parts fairly weak so I could see one of the forces I am trying to use.
-
My chain and sprockets have already failed in keeping things in sync. I knew they would not hold up for a long time but I thought that I would get a few weeks of testing before they went bad.
What I was playing with was making the motions and stuff that I wanted, not sure on the force values but some of the methods of interacting did work, again not sure on the force values.
-
so my primary test-bed has proven many times and in different ways that my larger diameter wheel can drive a smaller diameter wheel at a 1:1 rotation and a 1:1 force for a 1:1 work transfer, it goes the other way as well.
This is all accomplished by some simple but maybe a little clever design where the hidden distance is allowed to exist. Think about this way, if I have a 100mm radius turning an 80mm radius I would need some kind of gear that has that 5:4 ratio, mine is a virtual gear but it is real enough to perform the function needed.
My new test-part is doing the same kind of thing but using a real gear instead and I am trying to design it so that it is a poor gear design, I am willing to throw away some work to losses.
I am shooting for basically a 100mm diameter turning an 80mm diameter at a 1:1 rotation and with a 1:1 force but minus those losses I am designing into the system.
I am going this rout since my chain and sprockets are not doing so well.
-
To be a little clearer, what I think I can do is throw away the change in distance so that my input force on the 100mm is the same force out on the 80mm,,, say 1N of force in and 1N of force out at a 1:1 rotation.
-
Throw away is not really correct but it kind of shares what I am not overly concerned with.
So, my test-bed allows me to transfer the work, via the perimeter of the wheels, through the system at a 1:1 rotation and a 1:1 force as seen by each wheel. I use the same size drive disc on each wheel and measure the same force and the same change in angle.
None of that should be a surprise since our math requires that this be the case as far as final values go.
I am mentioning this again because the perimeters are not the same and so there is a difference of distance traveled, this difference is made up for with the work transfer function of the designed interaction. This WTF is another exchange\conversion that happens internally, something I now have access to.
-
I have a build error that I did not know about. I have 8 sub-systems with a part that connects to all of them and they all are supposed to be in sync with their motions and stuff, well somehow either a design error or a slicing error or a Freecad error, the systems are sort of indexed so that they all only line up when the parts are put together in one orientation and that still does not work nicely. The system is supposed to allow for a free rotation between all the parts but as the system rotates it starts to bind and that spring loads the actual parts so there are 4 positions that it wants to stop in. This is not a good thing for me.
This test-part was setup so that I could use it for testing 2 different thoughts on how to transfer the work.
One way would require that I flip the input and output functions at the appropriate time while maintaining a constant reactionary condition. I could see Bessler using something like this for his wheel since Gravity is directional it would provide a means to facilitate the flip and then he might be able to use the constant reactionary condition as his actual output.
Another way, which I am preferring to use, has a constant input and constant output and no flipping required. It needs a little better relational control and I think I should use a couple of limiters.
I have made my parts somewhat weak so that I can see the forces acting upon those parts, they bend and flex and it is this bending that I am using as a visual guide for the build. Obviously the final build parts will be stronger.
Another reason I chose to use 8 sub-systems is because each one can only pass a smallish amount of torque, 8 working together means I could pass 8 times as much before failure, hence the chain and sprockets to not only pass the output via the chain but the chain was supposed to keep all the parts in sync.
What I am seeing from my flawed test-bed, going for the constant setup, is that the forces and directions are doing what I was expecting. I have taken what should be an external rotation and stopped it and turned it into a virtual rotation.
I still need to design the feedback system that is needed between to parts, one moves a little relative and the other moves more, this work is the feedback I need to build for.
-
This should not need to be said, but, no engineer would design one of my systems to work the way I am using it
-
Not sure what this means.
While I am just playing and printing more parts for testing stuff.
I set things up in a strange almost backwards way, then I use Mr. Hand to hold the output still and then use the other Mr. Hand to apply force against the input arm. I then shift things so that it is close to where I think a force gain situation should be,,, now the funny thing happens. I move the arm and of course the output moves then I take Mr. Hand on the output and I increase the resistive force he is applying. Well, Mr. Hand on the arm does not feel that increase in force but several of the parts bend and flex more, as I would expect from the increase in force from the output.
-
Still just playing around.
I have been looking at what kind of exchanges I can have and what makes sense and what does not,,
I thought of something last night and then after thinking about it it seemed like it would have to end up being a net zero, well I printed a few parts off overnight and thought I would just go ahead and play with them, then I found that I made one hole to big for the screws I am using, anyway I just used a nut and bole and locked those parts together instead of allowing them to move on a bearing.
So my gears haves a 24mm radius boss, so a pulley if you will, I used the teeth to hold onto 8 rubber bands and held the other side to a non moving part, then I used 1 rubber band on a part that has a 100mm radius, the 1 rubber band would stretch and then it would move everything and stretch out the 8 rubber bands.
Not sure if it means anything but I thought this usage would not do anything.
-
I got curious so I measured the force on the 8 rubber bands when pulled to the "starting" position length, that force was 7N
The hub piece has a 24mm radius.
I have 2 positions I can measure for an output.
The lock pin has a radius of 80mm and has a force at the "starting" position of 4N
The other pull point has a radius of 94mm and has a pull force at the "starting" position of 2.4N
My spring scale has 0.4N marks
Edit to add:
The "output" points and the input hub rotate at a 1:1 ratio
-
I am running with only 1 set of 2 arms in this system and with what I have played with my expectations, if I were able to run a full cycle test, would be that the system would start at a net zero condition, ramp up to a maximum gain then down to a net zero then down to a maximum loss and back up to a net zero. I would expect that the maximum gain would match the maximum loss and show a complete cycle as a net zero.
If I were to add another set of 2 arms to the system then I would expect to be able to use the new set of arms as the transfer component instead of the loss cycle from the first set of arms, add more sets and I could be higher on the gain section. The more sets the higher on the gain slope the system could be maintained.
-
I can only run a static test, with that I setup where I thought the system should show a loss and it does show a loss but not as much as I thought it would be, so maybe the max loss part of the cycle is not where I think it should be.
-
I was just thinking to myself that what if the net zero to max gain and back to net zero is for less than 180 degrees and the net zero to max loss and back to net zero is for more than 180 degrees. This would allow for the loss to be not as much force but over a longer rotational moment.
If that is the case I can still add in enough systems to stay on the gain side.
-
Funny thing,, I thought I would print some purposefully made parts to improve on the ones I kind of just scabbed together.
Well now I know that my measurements are not so bad, with the "new and improved" parts it looks like it changed the functionality to a net zero,, that is tongue in cheek there, when I go back so do the readings.
-
Trying to figure out the actual transfer modality is frustrating.
With that, I thought to myself that if what I think might be able to happen within the system does I could then use a low force, aka 2 rubber bands, to check to see if the same force applied on the same radius at a 1:1 turn ratio can in fact manifest 2 different forces.
My sprockets can be setup for this simple test within the test-part. I set it up that way and used 1 rubber band on 1 sprocket and then used another rubber band on another sprocket in opposition to the first sprocket and each seeing a non-moving attachment point to the test-part.
My assumption was that if indeed the interaction with the system manifested 2 different reactionary forces that one of the rubber bands would stretch the other one out until the forces were balanced throughout the system.
I have ended up with one longer rubber band and one shorter rubber band and the system has a desire to rest in one position leaving the rubber bands in the long to short condition every time I move the system out of this resting position.
-
I tried the same thing on 2 sprockets next to each other, they are not as good I think but I tried it anyway.
So I still get the shift to a preferred position even tho it is not as strongly attracted to that position but here is the funny part, I thought to myself why not just hook the rubber bands directly to the sprockets with no fixed point of observation? Well when I do that and stretch out the rubber bands they do in fact move the system but they will move the system backwards compared to the fixed point direction and dependent on the system orientation.
-
right now as I am playing with this it is looking like it might be the slop that may be creating the motive values.
It might be that if I use the correct set of sprockets with there systems interacting and the slop is just right the system moves with some force. I have been trying the same relationship with other parts and other sprockets and I am finding that the shift is not all the same but that when there is one that the tendency is in the same direction.
-
I am focusing on the test I can do that has little to no involvement of Mr. Hand.
That test is setting the system in a particular spot, applying the rubber bands hooked on the teeth of the sprocket and a fixed point on the frame and setting the stretched length of the rubber bands the same.
-
I was getting inconstant tests with different sprockets, some showed no change others some change.
My frame is 2 pieces bolted together, there are many points that line up with these 2 parts BUT when I loosened the bolts and tweaked the alignment most of the "indexed" went away, unfortunately so did the reaction I was seeing from the one setup.
Conclusion: There was an internal storage and exchange that I did not understand and would most likely lead to a net zero condition.
Next step is to get back on the path I was heading, which I did and am again frustrated with issues.
What I see via a spreadsheet is that there is a potential to have a 1:1.5 relationship for work, however things load up, if you will, when trying to pass enough force to not worry about the frictional losses . So far the only thing I can say is that on one side it takes me approx. 1N to make the unloaded system move and it takes approx. 0.75N with the other side to make the system move. Same movement by the way since the sides are 1:1 that way.
-
made a few more parts.
I use the spring scale on one side with no load and the scale shows next to nothing to rotate, on the other it shows up to approx 0.25N to rotate so setup of things is closer.
I am using an arm for input and output and 2 spring scales.
I have used each scale on each arm.
The input measures approx. 1.5N and the output 1.75N.
I have run this a few times but since this is not a full rotation not much can be said other than this is less than what I thought I might see.
-
I am using a lot of bearings in this system. All those bearings are still packed with grease so they have some drag on them. I thought that I could just increase my input force and not worry about those losses so much.
I made a poor design choice\compromise in that my input interacts with a large group of levers and uses a bearing for that interaction, the bearing is supposed to run square with the lever face but my poor choice has a very small angle to the interaction and as I increase the force that angle forces the bearing to run across the face of the levers and it "pops" off of them. I increased the strength of the arm in an attempt to stop that but it still does it anyway.
Now I get to re-design all of those parts.
One interesting part to all of this is finding what WTF I need to break and redirect. I did not see where that was and am hoping that I have found the correct path to interact with. With that, I can see several ways of making that interaction if I am correct.
-
I am testing out a design change and have an interesting thing.
I have reduced the input\output radius, with that I can now apply approx. 15N of force on each side and nothing moves. If I want the system to rotate with this consideration in place I need to increase the force up to approx. 20N on whichever side I wish to be the driving side to move the other side at approx. 15N.
This could just be me not applying and moving things correctly but the same relationship appears to happen with more or less force being used.
I tried another one but the part was rather flexible and under force it would flex changing the rate of angular change side to side and showed a growth in force when I used the input side but the input side, due to the flex, moved a larger angle than the output side.
-
Just throwing this out there.
Using this design setup to cause the system to move, on the output side I need to apply approx. 1.5N of pull then to use the input side to move the system I need to apply approx. 1N of pull.
-
I have checked the mass and force on the drop weight, 2.5N. I have that connected to a string and am running that string over a pulley so it applies its force horizontal into the system. I have taken that pulley and used the other spring scale on the other side and it shows the same 2.5N even when I am raising and lowering the drop weight.
Moving slowly I have to apply from 3.2N to 3.6N of force to raise the drop weight, this is when the drop weight is on the output side. It is almost the same for the other side, but there are some spots where the force required jumps up to almost 7N when a part transitions its orientation.
-
I have made some parts strong enough so that I can actually see what the parts look like as they are interacting, those are the input arm and the interactive drive unit. This allows me to see what the lever system is doing as the forces are being applied and transferred from one area to another.
I have a pivot shaft that is supposed to be rigid but even with a small amount of force being transferred those rigid pivots are bending.
I am aware of these forces but I did not appreciate the reality of how large they become so I need to constrain that reactionary force issue before I can proceed.
-
simple solutions are good.
I am however having significant issues with the plastic parts deforming and not providing for a smooth interaction, when I use force the parts are almost grating by each other instead of sliding and rolling smoothly. I have one small spot where the input force to start things moving is less than 2.4N but then things get sticky and it takes over 4N to move but then the system kind of shoots forward and stops again until I pull further on the input.
It would seem that I need to use hard surfaces instead of plastic ones.
-
so my chain and sprockets are adding more losses than I thought. If I only use them to transfer between 2 sprockets I have an increase from about 0.5N up to about 3.5N. To take up all the slack and just start the motion of transfer is close to the drop weight value but to get the system to move more than that takes initially a lot more and then it drops down while motion is maintained,, somewhat due to Mr. Hand not being really good for this.
I took all that off and used one of my alignment jigs instead of the chain and things were interesting. I increased the drop weight to 4.5N and, depending on where in the rotation I measure, the input force goes from about 3.2N up to over 6N.
Watching the test-part bend and twist with all this is a bit of an eye opener as to how much constraint is going to be required.
-
sidetrack:
I have my multi-systems interacting with the input, so named because that is the way I have been viewing it to distinguish direction of throughput not as the ONLY way of using it.
My input and the multi are connected as such that one rotation of one drives one rotation of the other, they are a 1:1 rotation.
In looking for alternative methods of adding a constraint to stop the massive deflection of shafts and stuff I can up with a way but it has an interesting issue. The direction of rotation of the multi can go either CW or CCW but the input always rotates in the same direction CCW.
In one view then they are a 1:1 but in the other they are a 2:1, that is I can rotate the multi in the opposite direction of the input and one rotation is still one rotation which means that between them there are 2 revolutions.
This is not all new stuff, at least to me and I am guessing that it will end up leading to nowhere, but it is kind of fun.
-
I have drawn up something that is a little confusing.
I have a fixed input distance of 120mm
I have an angle of change at a constant 3.65 degrees for the input. I think this means my input work is going to be the same amount for any of the transfer system lengths.
the output is 45 degrees of rotation on a 10mm arm.
the longer I make the transfer system components the interactive angle between the input and output changes, ie: at one set of length values (468mm) the range goes from 106.01 to 74.83, many times longer (12m) the range goes from 115.8 to 71.37. This is in consideration that a 90 degree interaction angle would be the best force transfer.
I think that this could be a kind of non-rotating lever (NRL) in that my input distance is always the same.
-
not really useful for much but an interesting method.
-
I have rebuilt my test-part, I have made it smaller and with only 4 subsystems.
They all are mounted in bearings on a frame and they are all connected to each other with a single common part. They are all connected so that there rotational change is the same and in the same direction as each other.
I have test pulleys that are 31mm radius and are mounted on each subsystem shaft. I am using these to test input and output stuff so that I can identify and follow what the feedback systems are within the whole system.
On one pulley I have wrapped a string with a 180g mass hanging on it, well I can not find my other 180g mass right now so on another pulley that is opposite the first one I have wrapped a string with a 250g mass hanging on it.
The drop and lift distances are the same due to the pulley radius being the same and the angle change being the same.
There are force changes within the systems and in one point of view that can be viewed as a 3:1 and this is where the feedback part is that I am looking into, it should cancel that advantage.
So,, I start the first test with the 180g mass low and the 250g mass high, let the system slip through my fingers and the 250g goes down and the 180g goes up nicely for a 180 degree rotation. Now I reset my system for the other 180 degrees of rotation but I set the 180g mass high and the 250g mass low.
I use Mr. Finger to help the 180g mass to move down, a small amount of rotation and the 250g mass will drop back down, go past that value and the 250g mass no longer can lift the 180g mass, Mr. Finger is noticing a "heavy sluggish" feel but it is the same for either direction of rotation, keep rotating and Mr. Finger notices this "feeling" is reducing until there is a point where it is very easy to rotate the system in either direction a little bit, 10-20 degrees.
I tried to add 70g to the 180g mass to see if it would lift the 250g mass, well it did not work so well and the sluggish feeling got much worse.
I have to improve a few parts and then make sure that I have not created an artificial force potential with the setup, nothing like loading a spring without realizing it to mess with things.
-
making parts and playing with things,, fun stuff.
So I have made this one setup that actually has a few rough spots but for the most part moves OK, a little resistance but it rotates.
I found 2 electric motors that weigh 80g and am using those for my drop weights.
In the setup I am playing with right now I have 4 pulleys that go round and round but do not rotate :) Roberval anyone.
I have my 2 drop weights on opposite pulleys so the system should be balanced but it is not, both weights move the same distance, one down the other up and it is always the same weight, within this range of rotation, that will fall with extra force raising the other weight.
I actually was expecting either nothing, or the opposite of what I have so this is getting interesting.
-
just as a note; I am using gravity as a source, BUT, I have to supply both the motive force and the force of opposition, both of these are courtesy of gravity.
-
another note; a belt and pulley or a chain and sprocket are one-way force transfer devices. You can not transfer any force on the slack side of the belt\chain. You can MAKE and external constraint system to allow you to transfer a push for either but by themselves they can only transfer a pull.
-
I can add any number of sources to the slack side so long as the added force into the belt\chain is in the same direction, it would not be a source if it were not in the same direction.
To make or impart any FORCE there must be a force of opposition, an output if you will, or a sink. This point of opposition then determines the tension side and slack side and if the system were closed upon itself, like a belt or chain usually is, then I can instantly change where the slack and tension sides are, source and sink.
I can also have any number of outputs ,and as it should be the output(s) must equal the input(s). Any force potential that is not used by one of the outputs is passed through to another output that may be at any point along the path of the belt\chain.
Since force does not have a velocity or a distance then the modality of how the force is manifested does not matter. In the real world you just can not have your sources run through, physically, any other part.
So the belt\chain then, in a sense, becomes a storage medium for potential similar to momentum and how that potential is depleted is how the work is performed.
-
This is a little petty maybe but I want my test-part to use a different storage potential than the test-bed. The test-bed uses CP as the storage medium for work transfer and it looks like I can do something similar with the test-part but I want it to do things differently.
If I look at the "how" that may work and then look at what I thought the test-bed was doing,, well I might of over-thought the "how" for the test-bed.
-
On a 30mm arm I have hanging a 185g mass, on a 60mm arm I have hanging a 90g mass through a certain area of rotation the 90g mass picks up the 185g mass with ease, I need to add 30g to the 185g to balance things out. I changed my applied points so that they are 180 degrees across the same axle and on the same lever arm.
The complete lever arm weighs 30g and enough of the arm and parts are on the 90g side that when I weigh that side it is 20g. So this should then be close enough to being balanced. 30@210 60@110 roughly speaking is very close.
I have a range where this seems to hold and a range where it does not.
The whole system is very stiff to move but the weights on the arms move the whole system easy, so that is the internal moving parts are very stiff but the whole system is very free to rotate and move with no masses.
I am looking into a funny thing with this part of rotation relative to the direction of gravity and the relative positions of my system.
-
Need to add that I have a means of adding in an internal potential, I am using rubber bands to add that potential.
I have been able to add the correct number of rubber bands so that when I hang the 90g mass on the 30mm side I can almost balance the arm, remember that the other side of the arm weighs approx. 20g so not much.
Within the range the internal potential holds, outside the range it does not.
I took the whole system off the mount and and placed the central shaft in a set of bearings and with the rubber bands in place I needed to add 45g to the 30mm side to bring that lever level, so the whole system is balanced around the axle for the lever.
I just did something silly, I mounted the bearing block in my vise and then put the lever axle into that bearing block. My spring scale weighs approx. 45g in total, the scale shows 30g when I hang it from the hook by itself. So for 180 degrees of rotation that spring scale is almost enough weight to hold the lever level, (the scale shows 30g while hanging by itself), pull on it so about about 35g is showing and this comes to a peak, so from no extra needed up to about 5g extra and then back to no extra needed. The other 180 degrees of rotation and I need to pick up on that spring scale a little bit, down to approx 25 and this is again from none needed up to a max and then back down to none needed.
-
This could be fun,,
I am printing out some test parts for the test-part that on paper shows a 1:1.314 in to out
I used pulleys on the first re-do and now I am using gears, making all those belts was a bother cause it is hard to get the tpu to fuse straight, and then to get the correct tension in it all.
-
Made some quick parts for some quick tests.
I am using both gears and pulleys. The gears are 1:1 between systems but not to the go-between part, the pulleys are a 1:1 and connect all the gears together.
My input forces gears to move around the go-between part, that gear rotation is transferred to other gears which then rotate about the go-between part, the go-between part is the key.
Quick tests are looking very positive until the belt slips,, very annoying since now I am going to have to redesign things so that does not happen.
-
I will say that if I ignore the go-between part the math holds true that the if input side is a value of 1 then the output side is a value of 1.314, that is just mechanics at work.
So I am needing to test for the go-between part, which by itself is performing no added work, it is simply transferring the input to the output like a belt does for a pulley, or an idler gear,, and the like.
When the belt is not slipping it looks possible but when trying to load the belt slips and then no results to be had.
-
While I am trying to figure out how to design things I tried something silly and it has me a little confused.
I have holes in the pulleys so I can secure a string, I ran that string across the system to the other pulley and out through that hole so now I have a limited motion of no slipping.
I thought that I could use a weight for the input and output comparison, and I thought that it should go one side heavier and one side lighter to balance.
The whole thing rotates around on a central shaft and when I have the weights equidistant from that shaft the weight value is opposite of what I thought. What I have right now is 60g going down on one side of the shaft to 90g going up on the other side @the same distance of drop,,, roughly speaking. Making the system shift either direction is actually rather easy and light feeling and it seems as though the path I choose the string to take has an impact as well.
I thought that the 90g going down should lift 90g and that to balance the forces I should add to the lift side where what I seem to have is that the lift side is taking less input,, maybe I am identifying the sides backwards??
-
just double checking things.
the lighter weight is 70g, the lighter weight is 50mm from the central shaft
the heavier weight is 90g, the heavier weight is 55mm from the central shaft
-
the systems are not equidistant from the shaft so I am thinking that it is the difference that is providing for this interaction, however to counter the one side adds a force into the whole system so I would need to counter each system relative to the central shaft.
-
I think I might of found the pathway that I need to interact with, it is not really the way I thought it would be.
I am printing off more slightly changed parts, I keep getting the slop and printer error and stuff a little off so I get into either a miss or a bind and keep making small changes to get each part to fit and work together, it was in going over all this so many times that I saw what might stop the system from doing what I wanted it to do, that is provide me with a 1:1.314 gain in work.
Now to print more parts and see what I see, then print more parts from there.
-
My setup needs some structural parts added but the functional parts are in place well enough for a few basic tests.
My spring scales are problematic in that I need to overload the system to get smooth readings and that drives the system out of sync so that a limiter I have in place comes into play and locks things. While it is in this condition if I pull on the output with 1200g I only need to supply 900g to the input, however if I pull on the input with 900g I only need to supply 1000g to 1050g on the output.
I changed up to dropping weights, each are 90g. When I reset so that the input and output are synced back up the input side has a very slight tendency to fall. I add another 30g to the output side and it has a small tendency to fall. If I give the system a spin it moves further one way more than the other, that is to say that the heavier side will continue to move down further than lifting it up. If I put the 30g on the input side it falls without needing to spin.
I increased the drop mass to 190g and the system seems to be balanced, no tendency to move either way. I add 30g to the output side and it drops fairly quickly, when I move the same 30g to the input side it drops very slowly and will stop moving until I give the system a little push.
Not the results I was hoping for, but not the results I was expecting. I was expecting the system to stay balanced and behave in a normal fashion.
-
I upped the weight to 270g and things appeared to become more "normal".
Then I thought about the limiter, so I shifted the system one way and not much change then I shifted it the other way and now the output side having the extra 30g goes up, very slowly and I need to nudge things so I am thinking that what I thought was "synced" may be in the wrong place.
-
trying all sorts of little changes and at this point the one thing I have noticed is that when using the 190g weights the system behaves a little differently than when I use more or less.
I add 30g to the output side and the nudge the system and it will slowly fall down a little bit, put it on the input side and I do not have to nudge the system at all. I have tried with my sync point moved, using CW and CCW rotation, switching the weights, so it seems that there is a preferred force direction of change, how much and why, not sure.
-
Found an issue with the build.
2 pulleys printed from the same printer and the same file gave me 2 slightly different diameters. These are then walking around one another so that is what I think is showing the results I have been getting.