Need help with a mechanism

[NathanR]

I'm working on a 1:110 scale model of the Apollo 11 launchpad (Currently over 6700 pieces, WIP thread here), but I need help creating a gearbox that lets me rotate the service arms out of the way using a couple of knobs on the back.

The top of the tower looks like this (LDD file here):

The first 8 service arms are all mounted on a single 82L axle running the height of the tower (pink), and will rotate 73 degrees into the wall facing the rocket.  It would be good to have a worm gear to help lock the arms in place, though I am aware the mechanism already has a lot of play in it because of the universal joint I used to offset the axle at two points in the tower. The last arm, where the crew access the rocket, is mounted separately and rotates over 180 degrees to the side wall (yellow axle).  

I know this looks like I haven't even attempted a solution myself, but I can't even see how to turn each axle 90 degrees with a pair of cogs in the space available.  Despite years of enjoyment from technic sets I've never managed to build any working/reliable mechanism from a box of cogs without an instruction manual to guide me (seriously, I still have no idea how anyone was able to design set 8002... and the Mercedes Arocs gearbox is just "magic happens here").  So any designs, advice, or suggestions would be much appreciated.

[Aventador2004]

Looks hard, kinda sloppy ATM would you allow me to move where the knobs are?

 

[Bublehead]

Could you live with one knob that turns both?  I think you could engineer a solution where one knob rotates the crew gantry arm 180 and then use a shorter connecting link to turn the other shaft a fraction of that 180, or possible do a 2 to 1 gear ratio to get 180 on one shaft and 90 degrees on the other.  Aventador2004 looks like he wants the challenge    BH

[NathanR]

43 minutes ago, Aventador2004 said:

Looks hard, kinda sloppy ATM would you allow me to move where the knobs are?

Thanks for looking at this! Feel free to move the control knobs, though my preference would be to keep them on that side of the tower.  

The entire top deck is quite rough at the moment - while largely accurate, I don't mind changing this level so the cogs can fit in.

5 minutes ago, Bublehead said:

Could you live with one knob that turns both?

Not really... the crew gantry got rotated out the way well before the others.  In fact some of the other access arms got rotated out the way early as well, but I know it's impossible to have co-axial rotation at this scale.

And actually, a 2:1 gear ratio wouldn't quite work.  The main arms are rotated 73 degrees from the tower, not 90.  The crew access arm is >180 degrees.

Edited May 10, 2018 by NathanR

[Aventador2004]

9 minutes ago, NathanR said:

Thanks for looking at this! Feel free to move the control knobs, though my preference would be to keep them on that side of the tower.  

Great, I will try some things.

[Bublehead]

Coaxial movement is possible at this scale, but not coaxial control. That would require the arms not be keyed at all to the main shaft and then you would have multiple parallel shafts that would have lever arms to apply motive force.  IIRC, there are "pipes" and liquid gas lines that run up the side of the tower... I would be tempted to use them as camouflage for my control shafts the I needed to control each gantry arm.  Just some food for thought.  BH

[Aventador2004]

Umm, Question: do you have the complete structure LXF, Just so I can see the entire connected part?

Can the Bulky arm be move half a stud?

[lcvisser]

I thought that on the real structure, the arms swing away passively, i.e. by gravity. Would that be easier to engineer? You'd need to have the axis of rotation under an angle, and then some simple lock that keeps it in place until released. Then your control needs to only release the lock.

[Aventador2004]

1 minute ago, Ludo Visser said:

I thought that on the real structure, the arms swing away passively, i.e. by gravity. Would that be easier to engineer? You'd need to have the axis of rotation under an angle, and then some simple lock that keeps it in place until released. Then your control needs to only release the lock.

I actually am getting it to work how it is, but that is a good idea.

[NathanR]

@Aventador2004

You can get the current complete launch tower (plus rocket) LDD file here: https://www.bricksafe.com/files/NathanR/real-space/apollo-lut/ApolloLUTv38.lxf

36 minutes ago, Aventador2004 said:

Can the Bulky arm be move half a stud?

Which arm do you mean? The one that looks like a giant A-frame is the damper arm, it raises up vertically (the vertical position arm is included in the LDD file). The arm with the white block at the end is the crew access arm, when you see it next to the rocket you'll realise why I don't want to move it... but I will if it makes the gear system easier.

 

10 minutes ago, Ludo Visser said:

I thought that on the real structure, the arms swing away passively, i.e. by gravity. Would that be easier to engineer? You'd need to have the axis of rotation under an angle, and then some simple lock that keeps it in place until released. Then your control needs to only release the lock.

You know I'm not sure.  I thought the arms swung back under motor power, not gravity, but there was a complex latchback mechanism to ensure the arm didn't bounce off the tower and hit the rocket during its ascent (causing it to explode). A gravity lock would be a good idea, but given the arms are rotating not falling I'm not sure how well that would work, wouldn't you need to tilt the arms slightly?

[Aventador2004]

@NathanR

Here is my work, hope it helps!

• Gear for Large Arm
• Pull, Push lever for walkway.

Lego Digital Design File

[pagicence]

@NathanR I have given some thought about your problem. According to your LXF file, the service arm has 73 degrees travel and the crew access arm has 192 degrees travel.  If you want a “gearbox” inside the space you have intendent to put it in, it’s no problem. I did some calculations and come to a conclusion that everything can be moved by one handle in the back where instead of 2 gear wheels. And you don’t even need a gearbox. It all can be done by just 2 gear wheels inside.

So 192:73=2.63, that means that you need a 1:2.63 output for crew access arm and a 1:1 output for service arm. The combination of 8 and 20 teeth gear wheels gives you the ratio of 2.5 which is pretty close to 2.63. That means that the service arm will rotate 73 degrees and crew access arm will rotate 182 degrees. I think you can live with that.

The lever which you will rotate in the back is directly connected to the pink axle with no reduction and on that axle will sit 20T gear, which will turn 8T gear in the opposite direction, which will be located on the separate axle connected to the yellow axle. That way both arms will start to turn at the same time and finish at the same time but in different speeds. Given the greater speed of the crew access arm it will actually get out of the way quicker and it will travel longer.

Or if you want it more precise you can use combination of 16T:8T+16T:12T. This way you will get 1:2.66 final gear ratio which is even closer to 2.63. That means that the service arm will rotate 72 degrees and the crew access arm will rotate 192 degrees. So placement of the gears is: 16T on pink axle, which will rotate 8T on separate (blue) axle, where will also sit 16T which will rotate 12T on yellow axle. Now in this case you have a gearbox.

When you’re making a 90 degrees connection with axles, you have to be careful with the axle turning direction in order to make the purple and yellow axle turn the opposite ways.

Maybe @Aventador2004 can figure it out how to place the gears and connect everything, because I have no idea. I’m god with theories but when it comes to execution… I get stuck pretty easy.

I don't know how and where to implement a worm gear, in my opinion that will cause too many bends and it will complicate things even more.

This is some rough placement of the gears.

Solution 1

Solution 2

Edited May 11, 2018 by pagicence

[NathanR]

@Aventador2004 Thanks for the file! I love the pull/push lever mechanism for the lower arms, but I think it's set up wrong - it seems to push the arm to 90 degrees away from the tower, instead of folding it into the wall. I will dig out some technic bricks and have a play later tonight.  

The top crew access arm has a nice simple gearing system, but the arm can't spin the full 180 degrees to rest against the tower wall, it collides with the 91501 2x2 corner wall element.  I will see if I can move the mechanism higher up the tower so I can leave the arm at it's original height.

@pagicence Wow... mind blown, I never thought about having the two mechanisms linked but that really does look fantastic... Just curious but is it safe/advisable to use universal joints to turn 90 degree corners?

[Aventador2004]

3 minutes ago, NathanR said:

@Aventador2004 Thanks for the file! I love the pull/push lever mechanism for the lower arms, but I think it's set up wrong - it seems to push the arm to 90 degrees away from the tower, instead of folding it into the wall. I will dig out some technic bricks and have a play later tonight.  The top crew access arm has a nice simple gearing system, but the arm can't spin the full 180 degrees to rest against the tower wall, it collides with the 91501 2x2 corner wall element.  I will see if I can move the mechanism higher up the tower so I can leave the arm at it's original height.

Oh, I mistook where they folded, sorry. Hope it provided a concept. Will have a second look at it.

[NathanR]

No worries, it's certainly given me a great starting point!

[pagicence]

1 hour ago, NathanR said:

Just curious but is it safe/advisable to use universal joints to turn 90 degree corners?

Theoretically maybe... I don't know. Practically hell no. You do it with 2 bevel gears. I can try to implement the first solution if you are interested.

[Aventador2004]

2 hours ago, NathanR said:

No worries, it's certainly given me a great starting point!

Great, My brother really would love to see it finished, he has the Saturn V.

[Captainowie]

3 hours ago, pagicence said:

Theoretically maybe... I don't know. Practically hell no.

That's partly true. With just one joint, it won't even work theoretically.

But you can use two U-joints end-to-end to move rotation 90 degrees just fine, and in situations where the axle needs to be right up against a wall through the whole of its length (i.e. no room for bevel gears) that's probably the best solution.

Owen.

[NathanR]

I think I have a solution for the lower arms:

LDD file here.  It seems to run smoothly enough in real bricks (which was a surprise for me, it's the first time a mechanism I built hasn't fallen apart when I looked at it).  I'm not sure how well it will work on the full tower given the weight of the arms on the axle being turned - what do you think?  And I still need to figure out how to disguise it, either above or below the next floor up the tower.

[pagicence]

I think that you should definitely put it just below the floor. It will be out of sight. I would suggest that you brace the pulling axle at one more place and to put that 9L beam below the axle (of course you will replace it with 7L beam, obviously), that way you can brace the axle with another technic brick directly from the celling. As for how well will it work on the full tower, you are just going to have to test that yourself.

Edited May 12, 2018 by pagicence

[lcvisser]

On 10-5-2018 at 8:28 PM, NathanR said:

You know I'm not sure.  I thought the arms swung back under motor power, not gravity, but there was a complex latchback mechanism to ensure the arm didn't bounce off the tower and hit the rocket during its ascent (causing it to explode). A gravity lock would be a good idea, but given the arms are rotating not falling I'm not sure how well that would work, wouldn't you need to tilt the arms slightly?

I know that I'm a bit late with this, but it was difficult to find information on this -- apparently the LUT was not the most glamorous part of the operation  After a while (long while...), I found this paper from way back and it seems that the umbilical arms were actuated by pneumatic pistons pushing the arms away from the vehicle.

I think it would look really nice if the arms would fall backwards after unlocking, instead of having to rotate a knob, but indeed you'd have to tilt the axis of rotation a bit, since you don't have pneumatic pistons to push. I'm not sure how feasible tilting would be... The other solutions suggested probably have a better chance 

[NathanR]

I've hit a problem with integrating the swing-arm mechanism into the tower structure - the main swing-arm axle collides with the A-shaped damper arm, so I had to offset the axle backwards by one pin-hole using a pair of universal joints.  However, universal joints aren't constant-velocity joints so the input and output axles have different rotation angles at any given point in the cycle (ok, they line up at 90 degrees but that's a special case).

I've worked out all the trigonometry for the pivot point positions to get me a new "in-system" lever configuration.  I just need one final piece of information, and for the life of me I cannot figure this out:     For an assembly of two universal joints, connected by a 2L axle, and used to move the axle one pin hole to the side (i.e. the equivalent of using two 8T gears to transfer power between neighbouring axles), when the input axle is rotated at 73 degrees, what is the angle of the output axle?

(Note: I've tried finding this using LDD. However, the hinge align tool won't allow this linkage to be assembled, so I'm not sure if the 2L axle is too short to make this work.  I have managed to build it manually, but I keep getting varying results between 68 and 70 degrees.  I'm not sure if this is just LDD responding to the natural play or looseness of the universal joints.  I also tried computing the angle from the equations given the wikipedia page for universal joints, but I'm not sure which ones apply... is this combination of two universal joints what they call a double Cardan shaft?)

[Captainowie]

8 hours ago, NathanR said:

However, universal joints aren't constant-velocity joints so the input and output axles have different rotation angles at any given point in the cycle (ok, they line up at 90 degrees but that's a special case).

That's true, but when you have two of them joined together like you have, you can arrange them so that differences cancel out: the input axle is at the same angle as the output angle, but the intermediate shaft is different (but you don't care about that one). When you join the two U-joints with the 2L axle, make sure that the slit in the connectors line up.

Hope this helps,
Owen.

[pagicence]

9 hours ago, NathanR said:

For an assembly of two universal joints, connected by a 2L axle, and used to move the axle one pin hole to the side (i.e. the equivalent of using two 8T gears to transfer power between neighbouring axles), when the input axle is rotated at 73 degrees, what is the angle of the output axle?

The angle is the same.

9 hours ago, NathanR said:

I keep getting varying results between 68 and 70 degrees.

You will always get variations, both in LDD and in real life, because Lego gears have some level of deviation. That is the loose space between 2 teeth of the 2 gear wheels spinning one another. The space is necessary so the teeth can actually go into each other. But you shouldn't get variations of degree of the rotation when using U-joints. You just need to watch and follow the direction of rotation.

9 hours ago, NathanR said:

is this combination of two universal joints what they call a double Cardan shaft?

Yes it is.

[NathanR]

@Captainowie, @pagicence, Wow, this is fantastic, thanks for the info.  I really wasn't expecting the input/output axles to match up, but I just rebuilt the joint in LDD after rotating the input/output axles before attaching the universal joints, and it all fits together perfectly.  I guess building the joint at 0 degrees and then rotating must have messed up the LDD angles.