Most Challenging Mechanism

[Tobi Wankenobi]

What is the most challenging mechanism you have ever built (or even failed to buid)?

What mechanism do you think is impossible to build with Lego Technic?

I like to start this thread as a collection of challengig mechanisms. You may get inspired to try to realize something which is precieved as "impossible". Or people might post solutions to "impossible" mechansims in this thread. 

[Stereo]

Baker valve gear.  Kinematically it's doable, the problems are downsizing it to something that would fit in a MOC, making it sturdy enough to transmit any force, and having enough travel to actuate a pneumatic switch.  To a lesser extent it was difficult to figure out how it even works, from 100 year old scanned documents.

All the individual components are amenable to being made from Lego, as it's just straight and L shaped bars, the problem is that they're all stacked - green inside light blue inside dark blue inside red - so the minimum thickness ends up around 4 studs before including mount points on either side.  

How it works is the pivot at the middle of the red bar is "fixed" for a given mode of operation, so the blue bars pivot around that point as the crank pushes front to back, which rotates the bell crank.  Having the red pivot left of the bell crank pivot makes the crank and valve go opposite directions, right of the bell crank pivot, they go the same direction, which lets you reverse the operation of the pistons. (or stop the engine by putting it in the middle)  When they're running in opposite directions, light blue and green overlap on the bottom of the bellcrank so potentially the whole stack is actually red, blue, lblue, yellow, green, yellow, lblue, blue, red.  On real steam engines they use a S-shaped light blue bar to help with this issue, and have a bit different vertical location of the joints.

Edited March 15, 2024 by Stereo

[mahjqa]

The mechanism to store energy in the springs and to release it at the right time was very fiddly. It had to be coupled to a steering axle that could withstand the weight of the car dropping onto the wheels continuously. Weight distribution was also very delicate. On top of that, it had to look good as well.

[Davidz90]

A working calendar.

The first obstacle was devising way to encode the length of the months. My "eureka" moment was using a days dial with 32 positions and a hand that skips 1-5 days at the end depending on the month. A 12-sided cam with 5 possible heights turned out to be quite doable and compact.

[jorgeopesi]

Whatever loader I have done... I made around seven I think... to do a perfect working between the bucket and the arm and to get a decent playability with realism it is seriously underrated.

[Fluwoeb]

I made a 3-speed + R transmission exclusively using bevel gears... that was fun and cool. 

[aeh5040]

Probably the Tower of Hanoi solver. Still not completely happy with it, as it's not really reliable enough to run at a show: https://youtu.be/MHBsQS34Tu8

Edited March 16, 2024 by aeh5040

[Tobi Wankenobi]

1 hour ago, aeh5040 said:

Probably the Tower of Hanoi solver. Still not completely happy with it, as it's not really reliable enough to run at a show: https://youtu.be/MHBsQS34Tu8

Wow. That is purely mechanical, correct? No Electronic/Software Control!?

I find shooting mechanisms very challenging. Especially if they have a "automatic" function. I built one for my Fire & Rescue Tank, and making the shooting mechanism working reliably was about the same challenge as building the rest of the vehicle. I think the challenge comes from the fact that there are high speeds, high forces in combination with very limited space. Several sub-mechanisms need to be combined in very small space (accelerating the round, feeding the next one, preventing others from doing the same, preventing the loaded one from rolling away)

 

[Davidz90]

2 hours ago, aeh5040 said:

Probably the Tower of Hanoi solver

This one is truly jaw-dropping. Absolutely brilliant engineering.

[nu_montag]

I expected some intense techniques here but was totally blown away by what I'm seeing so far. Wow.

On 3/15/2024 at 3:00 PM, Stereo said:

Baker valve gear....the minimum thickness ends up around 4 studs before including mount points on either side. 

It took me looking at a video of one in motion to kinda sorta half understand what's going on. A large version of this could be cool in a standing display of steam train mechanisms—and size wouldn't be as big a concern. I'd love to see someone pull that off! With half beams, you might be able to condense the width a bit, but you might still have issues given the narrowest pins you could use would have a stud sticking out.

On 3/15/2024 at 3:08 PM, mahjqa said:

[Cadmium] The mechanism to store energy in the springs and to release it at the right time was very fiddly. It had to be coupled to a steering axle that could withstand the weight of the car dropping onto the wheels continuously....On top of that, it had to look good as well.

Whoa—you definitely achieved both the aesthetic and functional goals here. The cinematography in that video was very nice too.

5 hours ago, aeh5040 said:

Tower of Hanoi solver

Like @Tobi Wankenobi, I'm wondering if this is fully mechanical. It looks like it based on the wide shots. If so, could you start it from any arbitrary stack of rings across the columns, or does it have to run in this sequence? Does it do the logic/math on the fly or does it run on a predetermined sequence like a punch card? One minor comment—I lost sight of the start/end stacks in the video. If you were to take it to a con, would it be possible to indicate the start/end points to the audience?

 

I've never made anything this complex, but I'm hoping to experiment with things more in the future. The most complex Technic-adjacent build I've made so far was a piano-playing robot that I posted on the Lego Mindstorms gallery way back when. The mechanisms were simple—one wheel to drive it and two fingers to play the black/white keys, all directly driven by NXT motors—but accounting for the motion in the timings and picking songs that sounded nice when played at speed presented a good programming challenging. I wish I'd followed through with the idea to add an IR link to some Power Functions that could depress the pedals...I would have enjoyed the challenge of locking it in place and delivering enough force for this.

[2GodBDGlory]

That's a good question, that makes for some good discussion! Here's a few projects of mine that come to mind:

1:7 Bugatti Chiron:

This thing was absurdly challenging... It had a huge number of complicated mechanisms all trying to coexist in the same shell, which made for major difficulties. I wouldn't say any single function was overwhelmingly complex, though many were fairly fancy, but just making them all work together was very hard!

 

Transforming Plymouth Superbird:

This one was very challenging because of the large number of folding surfaces that all had to move together, at the right times, and at the right speeds. Adding to the difficulty was that the model didn't have much volume in its transformed stage, so there really wasn't much available space!

 

Mechanical calculator:

This might be the closest to what you're looking for in this thread! This was one where I was spending days trying to think through the logic of the whole machine, and just trying to understand the theory of how it could work. Tricky stuff!

On 3/15/2024 at 4:08 PM, mahjqa said:

The mechanism to store energy in the springs and to release it at the right time was very fiddly. It had to be coupled to a steering axle that could withstand the weight of the car dropping onto the wheels continuously. Weight distribution was also very delicate. On top of that, it had to look good as well.

I hadn't seen that before! Looks interesting and challenging!

On 3/15/2024 at 7:04 PM, Fluwoeb said:

I made a 3-speed + R transmission exclusively using bevel gears... that was fun and cool. 

Were all the gears engaging via their bevels, or were there double-bevels meshing like spur gears? I'd be interested to see that!

[Fluwoeb]

16 hours ago, 2GodBDGlory said:

Were all the gears engaging via their bevels, or were there double-bevels meshing like spur gears? I'd be interested to see that!

I had all the bevel gears engaging using their bevels (or at 90-degree angles to each other).  It was very challenging and there was a lot of half-beam spacing. There was one location where I had two 12-tooth bevel gears on opposite ends of a 2-stud-long axle with a single half-beam in-between them.  It was pretty goofy and bulky looking. I think if I were to show you I would build it again... (which would probably end up in a slightly different result) because I only had a short video of it and that was a while ago and not filmed well at all.  The main premise of the design was my discovery of how this and this fit together nicely (though probably not entirely "legal"), so I used that combination instead of the standard 16-tooth clutch gears.

[2GodBDGlory]

4 hours ago, Fluwoeb said:

I had all the bevel gears engaging using their bevels (or at 90-degree angles to each other).  It was very challenging and there was a lot of half-beam spacing. There was one location where I had two 12-tooth bevel gears on opposite ends of a 2-stud-long axle with a single half-beam in-between them.  It was pretty goofy and bulky looking. I think if I were to show you I would build it again... (which would probably end up in a slightly different result) because I only had a short video of it and that was a while ago and not filmed well at all.  The main premise of the design was my discovery of how this and this fit together nicely (though probably not entirely "legal"), so I used that combination instead of the standard 16-tooth clutch gears.

Ok!

That combination of those two parts sounds interesting--I'll have to try it sometime

[aeh5040]

On 3/16/2024 at 11:32 PM, nu_montag said:

Like @Tobi Wankenobi, I'm wondering if this is fully mechanical. It looks like it based on the wide shots. If so, could you start it from any arbitrary stack of rings across the columns, or does it have to run in this sequence? Does it do the logic/math on the fly or does it run on a predetermined sequence like a punch card? One minor comment—I lost sight of the start/end stacks in the video. If you were to take it to a con, would it be possible to indicate the start/end points to the audience?

Yes, it's fully mechanical. There is just one motor, running continuously.

It would not work from an arbitrary arrangement of discs - only one on the shortest route between start and finish. However, it's not a predetermined sequence - it does "work it out" as it goes along. In principle it would work with 10 or any number of discs (although not practical with LEGO).

The three pegs have different colours. After it transfers all the discs to one, it moves on to the next, and so on forever.

There are quite a few more details in thread about v1. This was partly pneumatic, but used otherwise the same method.

https://www.eurobricks.com/forum/index.php?/forums/topic/168443-moc-mechanical-tower-of-hanoi-solver/

[msk6003]

Mine is

Twin motor excavator

All six functions of the excavator operate using only two motors. All drive shafts had to be routed without overlapping, and in the case of version 2, the lower part of excavator is quite dense.

Studless 10262 DB5

It retains the same size and all gimmicks as the original model, but has been made studless and added a working engine and steering system. In particular, the ejection seat and front machine gun were the most difficult.

Studless 8480 Space shuttle A/B model

Although the clutch gear and mixer drum were inevitably used, there was a personal constraint to reproduce model only with parts that were in production at the time, and while the bottom was made very thin with approximately 1 stud to accommodate the satellite, the axes and gears for the operation of the rear booster and landing gear were used. Of course it had to be durable at the same time. Also, after completing the A model space shuttle, I planned to build a B model submarine using only those parts. This part was also tricky too.

[Fluwoeb]

On 3/16/2024 at 11:50 PM, 2GodBDGlory said:

Were all the gears engaging via their bevels, or were there double-bevels meshing like spur gears? I'd be interested to see that!

Here is a picture of it from when I originally built it 5+ years ago as you can see from the date stamp...  The only non-bevel gears are the visible 8t and 24t gears.

Edited June 14, 2024 by Fluwoeb

[Auroralampinen]

 I saw this video in youtube and i thinked is this possible to build with lego pieces this might be the most difficultest mechanism to replicate in lego world

[Lipko]

Landing gear for my plane, but because the space was so limited, not that it was a complicated mechanism.

 

[tseary]

Auroralampinen, I agree that the ABENICS gears would be a challenge in Lego. I imagine the friction and backlash would be critical. To say nothing of the motor control...

[tseary]

Long ago I tried to build a mechanical oscillator using a pair of the old 2L clutches. Each clutch had a forward/reverse output, and each output was connected to the opposite changeover catch by a white clutch gear. In theory each output would toggle back and forth endlessly. I couldn't get it to work though; in those days the smooth-sliding 3L axle joiner didn't exist, so it took too much torque to switch over. I should try again...

[LegoTT]

19 hours ago, tseary said:

Long ago I tried to build [...] oscillator

I guess there was a sorta electrical version of this in the dog thing in the multi control set back in 93.

 

https://lego.brickinstructions.com/lego_instructions/set/8082/Universal_Set_with_two_otors

[tseary]

On 9/25/2024 at 4:23 PM, LegoTT said:

I guess there was a sorta electrical version of this in the dog thing in the multi control set back in 93.

That's cool, I hadn't seen that before. It's essentially doing the same function that I tried to implement. If I understand the instructions, the big motor drives the wheels and steering, and slowly toggles the micro motor. Then the micro motor toggles the big motor, and so on.

It's interesting that it's applied in an automaton. That seems to be where a lot of mechanical sequencing finds a use.

[dantheman12]

A 6 speed + reverse manual gearbox with a single pivot point gearstick - in 1:8 scale.