Besslers prime mover and its enabler.

[Roxaway59]

Fletcher I look forward to looking at the new sims and I hope Robinhood can do the ones you are asking for in Algodoo. I have never tried modifying script in Algodoo.

Graham

[Fletcher]

This is more of a demonstration, of the sort thing i can do with Algodoo, than to give you exactly what you want Fletcher.

It is a lot easier than i thought, so fire away, tell me exactly what it is you want me to do, and I'll be happy to oblige.

Make it as precise as you can, with how many, how heavy and how far you want the weights, to get us started, and then we can adjust to get exactly what you want.

Mornin RH .. thanks for that, good work .. unfortunately I can't read any detail in your outputs no matter what I tried to do - shudda gone to spec savers ..

Ok .. in my WM sims I have a program constraint to work around for better predictive accuracy ( you may not ) - it recognizes that all disks can be either disks or spheres etc - what this means is that it treats all disks as tho they are point masses for inertial purposes - as you know any uniform disk will have a MOI calculated at 0.707 length of the radius which is the position where half the mass is inside this MOI radius and half outside it - the upshot being that regardless of whether a 2 meter diameter disk weighs 1 kg, or 100 kgs, or 1,000 kgs, its resistance ( MOI ) to being accelerated will be the same ( they all accelerate at the same rate ) in the program ( which is obviously nonsense ) ..

So when predicting or working with RPM is important a practical workaround is required to 'adjust' disk MOI - so I made the 2 meter backing disk 0.1 kg mass and added 3 grey disks at the same radius ( 0.707 of radius = 1.0 meter ) 120 degs apart - I then made it so that these all added to 100 kg of mass for inertial purposes i.e. 100 / 3 = 33.333 kg each .. MOI inertia accelerating problem solved ..

Next I added 8 blue disks/wts ( 45 degs apart ) at a radius of 0.75 meters ( 3/4 radius ) to allow the exterior to be empty of weights ..

I also added 8 red disks/wts ( 45 degs apart ) at a radius of 0.25 meters ( 1/4 radius ) ..

Then in the 'properties' field for each weight I replaced the mass amount with an 'IF logical statement' - this looked to my inputs for the mass to use ( say Total 47 kg / 8 wts = 5.875 kg each etc ) when to the right ( RHS ) of the x-axis vertical, and 1 kg / 8 wts = 0.125 kg each when on the LHS ( going straight to 0.0 kg mass can sometimes be problematic because the program knows all physical objects must have a mass so I minimized it as a workaround ) - the red wts did the same except it was reversed so that when on the RHS of x-axis the reds substituted the lesser amount and on the LHS substituted the greater amount etc .. first I had to create some outputs that logged the wts x-position for referencing to ..

The formula I made was in the mass field for each weight and was generically like this, IF ( Output[85].y3 > 0 , Input[84] / 8 , Input[83] / 8 )

Output[85].y3 was for the x-position of a wt and Input[84] and Input[83] were the substitute mass values to use depending on the x-position of the red and blue wts at any time for CW rotation ..

This made a system where as T79 said it appears that weights 'teleport' vertically ( 2 simultaneous lifts ( 0.75 m - 0.25 m = 0.5 m ) totaling 1.0 meters ) into the new position so we always had a heavy bias of system Com/CoG to the RHS of the axle where it stayed without any drift to the LHS of the axle ( where it would have negative torque ) ..

Then I set the accuracy to around 100 fps so I could capture the animations without any problems, and began to adjust the mass Inputs until I got it to reach a minimum of 50 RPM in 1.5 rotations .. this was the mass needing to be teleported to achieve that acceleration from a standing start and get to 50 RPM in 1.5 turns ..

ETA .. n.b. starting from a standing start it should quickly increase the RPM and not just instantly be at RPM because things have inertia that the OOB acceleration ( gravity torque ) has to progressively overcome, to be realistic ..

* My laptop is 8 years old and has many sticky numbers and letters - one good reason for my many edits from time to time ..

.....................

FWIW - a first step can be to not have inner radius weight and just have the outer circle of weights turning their mass on and off when crossing the x-axis vertical as it appears you have done - then adjust to get a 2 meter disk to reach 50 RPM in 1.5 turns - if that works well add in the extra detail similar to my sim ..

Cheers -f

[Robinhood46]

There are some bits in all that, that i i'm not sure i can replicate, some bits because Algodoo doesn't allow it, or i don't know how to do it, and some bits because I'm not exactly sure what you mean.

So we'll take it slowly and see if we can get there.

You're right i have turned the mass of each individual weight on and off, as it crosses the vertical line, or more precisely i have reduced the mass to next to nothing, for the same reason as you do with WM, ie, the programme does strange things if set at zero. For arguments sake 0.00012 kg is zero.

The reason i did variations, and ran the simulation at 0.10 speed, and then at the end ran it at normal speed, is because i haven't got a clue how to establish how long it takes to accelerate to the 50rpm you want. I was rather hoping you were going to say "the one with x kg weights looks good". Lol.
Algodoo never seems to respect rpm settings on axles, which is why i generally set them at 4 rpm or 7/8 top end, it is extremely rare i run them faster than that.
I think Graham has a better understanding of some aspects of Algodoo than i have, notably the things it can and can't show, calculate and graph. I have never really looked into it, because i don't really trust it and generally use it just for visualising ideas and sharing them.

I think the technique for magically changing the weight's mass is the same. It's the same as works function on spread sheets or data bases "IF" (SI in French) If x is good, do this, if not, do that. Obviously what is good can be a multitude of things. In the case here IF a higher value than the vertical line, weigh this much, IF not weigh that much. Straight forward stuff. I like to add the change of colour so as to let me know that it is happening, and to help adjust positions.

You say you needed to add the 3 weights at 120° to get the sim to recognise different masses of the wheel. I have just tried running a copy of the first wheel (far left) placed above the original, to keep the references for the mass switch identical, with the mass of the wheel set at 50KG (50% of the original, and after 8 rotations it had gained 90° of rotation, and then it progressively increased it's acceleration, at a faster rate than the original.
Does this mean there is no need to put the three weights, or have i incorrectly misinterpreted something? If i add the weights and they are not need, will this make it a false comparison?

That's a lot of waffling to not say a lot.
I'll make a single wheel, with the weights as you have on yours, without the three additional masses on the wheel.
i'll give you all the distances and masses and we'll se where we go from there.

[Robinhood46]

Just to be sure i am going in the right direction Fletcher.
I fumbled around and found how to show the rads per second.
This is running at 0.25 speed to show the acceleration.
I let it run 1.5 turns and it shows just under 50 rpm, if my calculations are correct, at 1.75 turns, it's just over 50 rpm.
The reason i let it run continually is because i noticed it stabilised and stopped accelerating.

Are you happy with this method for finding 50 rpm at 1.5 turns? What sort of precision are you looking for? Are we going in the right direction?

[Robinhood46]

Just to be sure i am going in the right direction Fletcher.
I fumbled around and found how to show the rads per second.
This is running at 0.25 speed to show the acceleration.
I let it run 1.5 turns and it shows just under 50 rpm, if my calculations are correct, at 1.75 turns, it's just over 50 rpm.
The reason i let it run continually is because i noticed it stabilised and stopped accelerating.

Are you happy with this method for finding 50 rpm at 1.5 turns? What sort of precision are you looking for? Are we going in the right direction?

Fat lot of good if you don't put the video.

https://www.youtube.com/watch?v=T_hSb8nRjWM

[Fletcher]

Ta for that .. I would change colours too if WM allowed it, but either it doesn't or I don't know how to do it - it would be visually useful to have but I have never seen it - it is a plus for Algodoo ..

I think you should try to establish if like WM it treats disks also as potential spheres - just run comparisons of the same sized disk but say 1 kg mass, 50 kg mass, 100 kg mass etc, with just a few outer weights teleporting and see if they have the same or different acceleration rates - that will give you a basis for comparison of MOI to work from - if no different you'll need to add inertia MOI masses to the disk like I did in WM - if they are different it is treating it as a disk and not a point mass etc and you don't need the extra MOI disks it seems ..

Turning mass on and off with an IF statement is very useful and good to see Algodoo allows it also .. fwiw any logical statement like AND, or OR, can be used in a field ..

Graham or Shadow may be able to advise you on monitoring the RPM, time, and number of revolutions outputs etc in Algodoo - Grraham likes to use Radians but I prefer to stick to SI units we all understand - of course this thing is not motor driven ( just gravity driven ) - also on playback speed - the point is there should be some ability to see how the RPM and number of revolutions are increasing with outputs, so you can adjust the main inputs until you get what you are looking for ..

Graham and Shadow can post Algodoo sims - I think they embed them in a .zip file ( which is allowed on the forum ) and then they automatically unzip when you download and open them ..

Best -f

ETA .. it should not stabilize RPM because it is always in extreme overbalance where the system CoM/CoG is circulating but always to the RHS of the axle, therefore it always has gravity torque and it can't stabilize and must continue to accelerate - you can try increasing the accuracy setting to see if this makes a difference - it would be good if you could post files so that between you, and Graham, and Shadow you can help each other and improve sims going forward ..

[Robinhood46]

It's not really Algodoo can do it, it's more you can do it with Algodoo. When you make modifications in the script menu, Algodoo tells you that it no longer guarantees the correct function, or it can't guarantee you will be able to undo or reset, or something along those lines. I was rather hesitant when i first started playing around in the script menu, but it's well worth the risk.
From what i understand the worse thing that can happen is you lose your current work, and have to restart with a new page and default settings. It can also ground a few thousand aeroplanes, shut down supermarkets and cause hospitals, banks and factories to put everything on hold, but don't tell anyone i did it.

I don't know what to think of the difference between 100 kg wheel and 50 kg wheel, 90° after 8 revolutions is not a lot, but it is showing a difference.


So this video shows a wheel with weights like yours.
0.75 M from the centre and 0.25 M. They all have identical weights, all 16. Bear in mind, when you say "47 kg total / 8 wts = 5.875" you do realise that 8 x 5.875 never exists, because of the miracle that makes them weigh less on the way up? It doesn't change anything, unless you are calculating something with the 47 kg.
They all have 5.875 when they are red and 0.125 when they are green.
The mass of the wheel is 100 kg.
As you can see; if we accept my method of observing the acceleration, we only have 4.53 rads, which is 43 rpm. Graham probably works with rads, for the same reason i do, that is, we don't know how to change it to rpm. Apologies for this Graham, if it is a choice.

So we have a discrepancy.
I have left everything at default settings, for wind, friction and gravity, should i adjust one of these?

https://www.youtube.com/watch?v=6Kgf8bMoUHs

[Robinhood46]

FWIW it stabilises at 122 rpm.

[Fletcher]

So you don't need the extra 3 wts for MOI adjustment by the look of it i.e. it only recognizes it as a disk, and not a sphere which has a different MOI factor ( that's why WM treats it as a point mass IINM ) - No, there are no frictions at all - clean skin sim - best case scenario .. if you have any frictions it will reduce the RPM - same reason it may be stabilizing at 122 RPM - eventually it will because of the calculation accuracy but that's not important ..

Suggest you increase the individual weights until you hit 50 RPM in 1.5 turns .. and remove any frictions etc ..

Regardless they are fairly comparable ..

[Robinhood46]

I lied about wind and friction, unintentionally.
I hadn't noticed wind and buoyancy was off, so when i put friction of the parts to zero, it changed nothing. Still the same acceleration. Still the same stabilisation.
I'll have a play tomorrow to see how far i need to deviate to hit 50 rpm in 1.5 turns.
I think i have done the necessary to publish (share) my scenes, so hopefully, once admin has validated my registration, i will be able to share it here.

I'm off for some sleep now, it's 3 30 am here.

[Fletcher]

Just ran mine beyond the Pause at 1.5 turns ( hit Run again ) - 280 RPM at 20 seconds, and still climbing ..

.............

.. 0 to 50 RPM in 1+1/2 revolutions , in 3.6 seconds .. now that's quick ! ..

[Robinhood46]

and still climbing

Is it still climbing a cliff face, or just a gentle slope?

[Robinhood46]

.. 0 to 50 RPM in 1+1/2 revolutions , in 3.6 seconds .. now that's quick ! ..

I have kept the weights position and mass, the same and only modified the weight of the wheel.

50 KG wheel = 50 rpm after 1 and 1/4 revolutions and takes 3.01 seconds
60 ________ = ___________1 ____3/8 ___________________3.35
70 ________= ____________1 ____9/16 _________________ 3.82

67 ________= ____________1 ____1/2 ___________________3.65

My techniques are bit doubtful, but best i could come up with.
For reaching 50 rpm, i run the wheel from stopped at 10% speed, and stop it when rads = 5.24, or when very close, and select a taj before or a taj after, depending on exact figures.
For time taken, i run it numerous times at 100%, until i have 3 times within 10 seconds of each other, on a stopwatch, and average them out. Between 3 and 6 tries is generally ample to get this.
They are not perfect, but i tried my hardest to be as accurate as poss.

67 kg wheel coming in at 1.5 turns in 3.65 seconds is, i think we can say, exactly the same as yours at 3.6.
But is this not exactly what was to be expected?
How can a wheel accelerate to exactly the same speed, over exactly the same distance, and take anything other, than exactly the same time?

I think the only thing this tells us, is that the two program do not treat the wheel in the same way, hence the need to modify the mass to find the same outcome.

Can you try having 6, 8, 10 or 12 weights that give the mass to the wheel, to see if you can find the 3 different figures i did, before hitting the target?

There is still hope that Shadow , Graham and whoever else wants to play with it, can.
I get this when i try for the time being.

"Your account has not been activated yet, please come back later".

[Roxaway59]

Robinhood I will certainly have a try but I am not technically minded when it comes to script. I have been experimenting today with positive feedback and trying to work out what mechanisms can help turn the wheel when it is pushed. I will have a look at how to change script in Algodoo.

Graham

[SHADOW]

Désolé, ce type d'expérimentation dépasse mes compétences et je n'ai pas bien saisi ce que vous essayez de faire en modifiant les scripts.
Je n'ai pas la version pro Algodoo donc limité pour les options.
J'ai relu le fil depuis la première Sim de Fletcher, vous souhaitez mettre en évidence de manière virtuelle ce que Bessler aurait pu réaliser physiquement?

Sorry, this kind of experimentation is beyond my skill and I didn’t quite get what you’re trying to do by changing the scripts.
I do not have the pro version Algodoo so limited for options.
I have reread the thread since Fletcher’s first Sim, you want to highlight in a virtual way what Bessler could have achieved physically?