[WIP] Straddle Carrier

[Lasse D]

This topic is for discussing and provide criticism of the straddle carrier that I am working on. This is the current status:

As seen in the video, there are some rather big flaws: The wheels come out of alignment and start ripping the sides away. I do not have a good solution for this yet, so please come with your input and I will try out various solutions.

Using tracks would not be realistic, but it might be the most workable solution in LEGO. I really want to find a solution that has wheels mounted as the current prototype.

Other areas of improvement:

- Add fittings, such as the driver's cabin.

- Add drivers to make it easier to align with containers

- Further improve the spreader so that it can lift containers even when fairly misaligned

- Increase height so containers can be stacked to 3 in height

- Get the EV3 to work and be controlled using a website

Functions

The spreader opens and closes the hooks when lifted, so only a single motor is required to work it.

If you can't see the video, then here is a sequence showing how containers can be picked up:

And another sequence more close up:

The other functions are driving (the four middle wheels are powered) and steering (all 8 wheels steer using the mini turntables)

So there you have it. There is still a lot of work to be done on this model. Feel free to chime in with ideas - then I can try them out.

[Lakop]

Well done for getting to this stage. I think the mechanism needs gearing down. It all moves a bit too quickly and I would double those wheels up (if you have more). Make your structure a little bit taller so it can lift the containers comfortably and with a bit more freedom. its a very good idea and you have done well so far.

H

[BusterHaus]

Gears skip when they aren't braced enough, and the axles can change position. 90 degree transfers can be braced properly using this piece when space is an issue:

The build looks very interesting so far, I'm interested to see how it progresses. 

[pleegwat]

Looks impressive. I agree on the speed, though analog controls may help on that as well - bringing the vehicle up to speed more slowly will make it look much more natural even if the top speed stays the same.

Glancing at your photos, it looks like the wheels on either side are turning by the same amount, and wheels on both sides are always running at the same speed. This is probably a part of your cornering problems - I think you need to include a differential on the drive, and some form of ackerman steering. That won't solve the entire problem (you're almost certainly have gear slack problems as well) but it should help.

[trekman]

This is a highly impressive build, and the result to me is quite good, although to others it might seem disappointing.  It has already been suggested about differentials in the driveline, but i wonder should you have separate motors for each side, as there is bound to be a lot of backlash in the driveline, which is not helping with your problems either.  If you had seperate drive motors for each side, then you would eliminate the need for differentials.  Also if you drove the outer wheels on each side, instead of the inner ones would that make any difference?

You likely need more bracing at the lowest point which does not restrict function.

This is one of the situations where one is recreating a vehicle where in real life there would be hydraulic motors.  They are used because the wheel drive speed could be separately managed.  I just fear that the best result would be nxt, ev3 or PU motors on each side which could be configured to suit the steering arc which the machine is following.

But well done so far.  I will follow this...

[pleegwat]

Bracing between the sides at the bottom is probably functionally impossible. Additional bracing at the top (Box over the cover, and diagonal braces) will probably help keep the structure more sturdy. Though I'd say addressing the driveline is more important.

[Bublehead]

Take a look at mine for some inspiration?

 

[Lasse D]

Thanks for the replies everyone! It is great to get so much input. Hopefully the final model will work really nicely.

The first issue to address is steering and how the wheels tend to pull to the sides as shown in the video. I have created a new wheel design which is 2 units lower. Here the new design is shown next to the initial prototype (which is same height as the yellow ones:

My apologies for the video quality. It will be two weeks before I can get back to my camera and proper lighting again.

The lowered height will hopefully reduce the amount of sideways wandering.

 

 

Another issue was gear slipping. This only happened between the mini turntables and the bevel gears. Unfortunately there is no way to brace this. One could use additional bevel gears, but that would require additional bulk with gears. I will instead try with a liftarms-based solution, which is why there is a pin in the upper right corner of the picture.

Using pins and liftarms should also allow for Ackerman steering geometry similarly to on my Model Team trucks as @pleegwat mentions.

Finally regarding steering. The new wheel design no longer uses the white gears. They acted as primitive differentials. With this new design I will have to add an open diff.

 

My attempts at programming the EV3 to receive commands over bluetooth is still a work in progress. The motions when powered using smart bricks will be less jerky. For the video I hooked up some PF controllers in order to show the functions - these will be removed once I get the EV3 up and working.

My next update will most likely be for bluetooth communications since I will not only be away from my camera, but also the LEGO for the next two weeks.

[Saberwing40k]

That looks loads better, but I think any issue you are going to have is the frame itself flexing. With that in mind, have you considered splitting the difference between this and a tracked version, and made a wheeled version that was skid steered? That might be easier to make strong, but issues with turning could make that a problem.

[Lasse D]

I have built a second prototype with an updated driveline and the new shorter wheels (but missing the drive motors and mechanics on top):

By using liftarms instead of gears, the wheels no longer fall out of alignment, and it has allowed me to add Ackerman steering geometry as well. By reducing the driven wheels from 4 to 2 enough space was freed up to allow for the new steering mechanism:

However. The big issue of having the legs pull to the sides remains. For this reason I have added the long gray liftarms. The theory is that legs pulling to the sides will pull against the liftarms. By having them connected to the steering, the wheels should steer the legs back under the straddle carrier.

Unfortunately the liftarms do not pull enough.

My next attempt will be to have the top extend down further on the sides, allowing the "error correcting" mechanism to provide more input than in this initial setup.

If this keeps on failing, I will look into "splitting the difference" as @Saberwing40k points it out. I will, however, first try to build some Mecanum wheels and see if their sideways movement will be sufficient in having the legs straighten out during normal operation.

[MajklSpajkl]

Hi @Lasse D, great project! I kinda overlooked your topic until today. Perhaps I can share my experience, as I built one back in 2018, but never made a presentation cause it was a failure in my eyes and the reason was driving and steering. I was super happy with the crane part. It was built for the scale of Mack truck set and containers of that cross section as used there (also my Kenworth truck) You can see a couple of photos of mine here and there is also one in my too-big signature photo :-) Maybe one day I will show more of it in a separate topic... I used ropes to lift the crane and the containers were grabbed from front and back, so different approach on the crane as yours - as far as the drivetrain and steering goes, you are on a similar bus as I was, but I think you still have a chance to change the destination to Success and not to Failure, where mine ended :-) So to explain why mine was a failure and why I believe you could succeed, should you make some changes...

1. Driving steered wheels:

In my opinion those two together are not possible with LEGO. The problem is, that there is always some slack in the steering mechanism from top table to the bottom wheels. And at least in my case, when a wheel was driven, it always had tendency to rotate a bit as well, which resulted in stretching the two sides apart or shrinking them together. So, I believe a 6-wheeler is a better (but way less cooler) option with 1st and 3rd wheel steered and only the second driven, but fixed. Left and right driven wheels should be connected with a differential in order to compensate different trajectories, they are running on (had those on mine). With second wheel only driven, driving won't affect the distance between both sides. I had a problem even when driving straight as the wheels had a tendency to rotate just because of torque through the turntables. I didn't want to give up on the 8 wheels setup and couldn't find the solution for that problem... I hope you do, but as said – 6-wheeler has better chances IMHO.

2. Steering:

I see you have left and right side of steering connected to one motor. I don't know if you considered that, inner side of the carrier should turn more that the outer side due to different radius of the trajectories of the wheels and when you turn the other way it goes vice versa. You need a sort of Ackerman steering. And to avoid lateral stretching or shrinking both sides need to have the same centre of the steering radius and front/rear wheels need to turn exactly enough less then wheels on 2nd and 3rd axis in order to keep the correct radius. While I kinda got correct difference between rotation of axles 1+4 vs 2+3, I wasn't able to find any mechanical solution for the Ackerman, so I decided to steer left and right side separately, each with its own motor and tried to visually adjust turning angles to get the proper radii of inner and outer wheels. Why I think you could do better than me? Mindstorms! Unfortunately, I don't have it, but I guess it is possible, with some math behind it - that you program proportional steering for left motor and right motor so that the turning radii would be in the correct relation between inner and outer side of the turn. I believe also the real machine has to have something like that, because each wheel is separately rotated. Also, if I am not mistaken the real machine has driving motors in the wheels so there is no drivetrain through the turntables and there is no effect of torque on the turn table.

So, I will keep an eye on your progress here, as I really would love to see you succeed - I think straddle carriers are awesome piece of machinery so I really am cheering for you.

Best regards,

Miha

Edited November 18, 2020 by MajklSpajkl

[pleegwat]

19 hours ago, MajklSpajkl said:

Mindstorms! Unfortunately, I don't have it, but I guess it is possible, with some math behind it - that you program proportional steering for left motor and right motor so that the turning radii would be in the correct relation between inner and outer side of the turn. I believe also the real machine has to have something like that, because each wheel is separately rotated.

I'd hazard they automatically compensate for the sides drifting away or toward each other as well. It may be possible to set up a mindstorm sensor to detect the skew?

[Mr Ogel]

Hi Lasse, beside the other points like steering or gearing down the lifting mechanism I would like to mention that rigidity is a main point as you experienced. If you look to the top part of a real straddle carriers it is not flat but similar thick to the supports. In your model it is only one liftarm thick but you can easily make it 5 layers high also giving you enough space for diagonal connections between supports and top. As you know liftarms are anyway not as rigid as studded technic bricks and can bend much easier, This makes it also more demanding especially for such a construction.

I liked also the short lfting of the carrier itself which of course does not happen in reality. In the end you already created a nice construction and you took a good challenge.

[Brickend]

Love the way this looks and an interesting problem!

I have a theory which may well be incorrect, but I wondered if the bending action is due to a torque reaction from the vertical driveshafts, applied over a long lever (legs)? This is made worse by steering due to the increased load on the driveshafts?

Stalling the drive train at a point just before the wheel stations without a driven wheel connected may prove or, more likely, disprove my theory. 

 

[Lasse D]

On 11/18/2020 at 10:50 PM, MajklSpajkl said:

Hi @Lasse D, great project! I kinda overlooked your topic until today. Perhaps I can share my experience, as I built one back in 2018, but never made a presentation cause it was a failure in my eyes and the reason was driving and steering. I was super happy with the crane part. It was built for the scale of Mack truck set and containers of that cross section as used there (also my Kenworth truck) You can see a couple of photos of mine here and there is also one in my too-big signature photo :-) Maybe one day I will show more of it in a separate topic... I used ropes to lift the crane and the containers were grabbed from front and back, so different approach on the crane as yours - as far as the drivetrain and steering goes, you are on a similar bus as I was, but I think you still have a chance to change the destination to Success and not to Failure, where mine ended :-) So to explain why mine was a failure and why I believe you could succeed, should you make some changes...

1. Driving steered wheels:

In my opinion those two together are not possible with LEGO. The problem is, that there is always some slack in the steering mechanism from top table to the bottom wheels. And at least in my case, when a wheel was driven, it always had tendency to rotate a bit as well, which resulted in stretching the two sides apart or shrinking them together. So, I believe a 6-wheeler is a better (but way less cooler) option with 1st and 3rd wheel steered and only the second driven, but fixed. Left and right driven wheels should be connected with a differential in order to compensate different trajectories, they are running on (had those on mine). With second wheel only driven, driving won't affect the distance between both sides. I had a problem even when driving straight as the wheels had a tendency to rotate just because of torque through the turntables. I didn't want to give up on the 8 wheels setup and couldn't find the solution for that problem... I hope you do, but as said – 6-wheeler has better chances IMHO.

2. Steering:

I see you have left and right side of steering connected to one motor. I don't know if you considered that, inner side of the carrier should turn more that the outer side due to different radius of the trajectories of the wheels and when you turn the other way it goes vice versa. You need a sort of Ackerman steering. And to avoid lateral stretching or shrinking both sides need to have the same centre of the steering radius and front/rear wheels need to turn exactly enough less then wheels on 2nd and 3rd axis in order to keep the correct radius. While I kinda got correct difference between rotation of axles 1+4 vs 2+3, I wasn't able to find any mechanical solution for the Ackerman, so I decided to steer left and right side separately, each with its own motor and tried to visually adjust turning angles to get the proper radii of inner and outer wheels. Why I think you could do better than me? Mindstorms! Unfortunately, I don't have it, but I guess it is possible, with some math behind it - that you program proportional steering for left motor and right motor so that the turning radii would be in the correct relation between inner and outer side of the turn. I believe also the real machine has to have something like that, because each wheel is separately rotated. Also, if I am not mistaken the real machine has driving motors in the wheels so there is no drivetrain through the turntables and there is no effect of torque on the turn table.

So, I will keep an eye on your progress here, as I really would love to see you succeed - I think straddle carriers are awesome piece of machinery so I really am cheering for you.

Best regards,

Miha

Holy moly Miha! Your straddle carrier looks amazing! You are clearly going for an ultra-realistic look with the lifting mechanism, how the spreader is able to pull out and just overall aesthetics. It pains me that it did not succeed. I have made some tests, and I have bad news. The slack in LEGO pieces makes the legs stray to the sides or inwards even when the steering is locked in the straight position! For this reason active steering adjustment is necessary, and not even a 6-wheeled approach with single-purpose wheels will work. @pleegwathas mentioned using Mindstorms, but it is not that easy to detect and account for the legs straying to the sides. My latest attempt is the following:

I am using the gray frames to "error-correct" the steering. The idea is that when the legs stray outward, the gray frames (not being connected to the legs) will pull the steering so that the wheels turn inward. Similarly they will correct to the other side when the legs go inward.

Unfortunately the steering is too heavy, and this approach does not work either:

The issue seems to be the complexity of the steering. The rotating sections in the middle cause Ackerman steering geometry to be introduced, but they also seem to be cause too much slack and resistance.

I will give it a last try with an even simpler steering mechanism.

 

On 11/25/2020 at 9:44 PM, Mr Ogel said:

Hi Lasse, beside the other points like steering or gearing down the lifting mechanism I would like to mention that rigidity is a main point as you experienced. If you look to the top part of a real straddle carriers it is not flat but similar thick to the supports. In your model it is only one liftarm thick but you can easily make it 5 layers high also giving you enough space for diagonal connections between supports and top. As you know liftarms are anyway not as rigid as studded technic bricks and can bend much easier, This makes it also more demanding especially for such a construction.

I liked also the short lfting of the carrier itself which of course does not happen in reality. In the end you already created a nice construction and you took a good challenge.

I'm glad that I am not the only one who got that idea! I thought the same, and was about to build a super rigid top section. However. It is the spreader which causes the main rigidity of the legs, since it is far lower placed. Unfortunately the soft nature of the liftarms makes the whole structure very flexible even with the spreader in the lowest position!

 

On 11/25/2020 at 10:54 PM, Brickend said:

Love the way this looks and an interesting problem!

I have a theory which may well be incorrect, but I wondered if the bending action is due to a torque reaction from the vertical driveshafts, applied over a long lever (legs)? This is made worse by steering due to the increased load on the driveshafts?

Stalling the drive train at a point just before the wheel stations without a driven wheel connected may prove or, more likely, disprove my theory. 

 

I have tested your theory by locking the steering and all the wheels were in the neutral/straight position. Even like this, the legs strayed to the sides as seen in the first picture. Having an ultra-rigid superstructure might help with this, but then we would be talking about something like 9 studs wide solid legs. I am instead looking at another alternate approach:

I know this throws realism out of the window, but I would like to see if I can make the model work so that it is fun to play with.

[MajklSpajkl]

Thanks Lasse, for your kind words - indeed the slack is the problem - My carrier was quite sturdy at the top but the spreader at the most bottom also didn't help. The main killer was the slack in the small turntables I think. Each individual wheel had just to mach free movement available.

I don't quite understand that solution with the gray frames pulling the steering in and out - was this statical solution or active with some motor or lever that does the job?

It might work on treads better yes, but takes all the realism out of the creation as you wrote. Still, I think it's worth a shot. Good luck.

[Lasse D]

I am explaining how it is supposed to work at 3:01 in this video:

In the video I also show the tracked version in action.

The major change from the wheeled version is fo course...

The NXT motors fit in nicely. I'm sure PF and PU motors would do too. The reason I am using NXT motors is that I want to make it "smart" and either be able to pick up containers automatically, or be remote controlled using Bluetooth devices. For now it is powered by PF since it is super easy and quick to implement for testing.

The spreader has been updated:

The red liftarm is the crucial part that makes the mechanism work. I causes the "clicking" mechanism that makes the fingers in the corners lock and unlock. I will make a separate post with how this works.

The lifting mechanism is super simple:

The white wheels cause the spreader to stop in top and bottom positions without destroying the parts.

Now I will look at that new simpler error-correcting steering mechanism for the red wheeled version.

[MajklSpajkl]

1 hour ago, Lasse D said:

I am explaining how it is supposed to work at 3:01 in this video:

Oh now I get it. Very clever idea, to bad it does not work. Very good solution for the steering geometry for different angles when turning left or right. 

The tracked version looks really playable and very fun.

I remain cheering for you to succeed on the red one ;-)

Best regards,

Miha

[vascolp]

Hi, very cool stuff!

You could try wheels without rubber tires.... just the rims or fake tires made with small link chains, for instance. Not that nice but it would probably slip enough to avoid leg spreading? 

About container decoration, you could make a decoration of 3x3 containers size, so you would have a puzzle that you would build using your machine!

[Mr Ogel]

Hi, nice to see the yellow version working, that's what Lego Technic is for. The steering of the red one works well - nicely done. 

But still back to the point that you need a rigid construction on the top of the carrier. The bridge carrying the container is in real life not made to hold the carrier itself together, it will just carry the container hold by the ropes connected to the winches on the top of the carrier. The point is that the force caused by the weight of the bridge and container is compressing the top and therefor bending the pillars - you can compare it to a beam bridge- would you like to drive such a bridge with a car when the pillars are bending? The yellow one is forwarding to force to the ground better since the track is also directly in the center of the pillars. 

You need to make the top and the pillars much more rigid, your grey liftarms are a first step on a longer way.

See a real life description of such a beam bridge: https://science.howstuffworks.com/engineering/civil/bridge3.htm

A nice challenge - stay on it and good luck!

 

[Lasse D]

On 12/29/2020 at 4:12 AM, vascolp said:

Hi, very cool stuff!

You could try wheels without rubber tires.... just the rims or fake tires made with small link chains, for instance. Not that nice but it would probably slip enough to avoid leg spreading? 

About container decoration, you could make a decoration of 3x3 containers size, so you would have a puzzle that you would build using your machine!

It is a good idea with hard tires in order to mitigate the issue. I have not found a pretty solution yet, but I like the idea of using chain links... They would be like reverse snow chains!

 

On 12/30/2020 at 8:56 PM, Mr Ogel said:

Hi, nice to see the yellow version working, that's what Lego Technic is for. The steering of the red one works well - nicely done. 

But still back to the point that you need a rigid construction on the top of the carrier. The bridge carrying the container is in real life not made to hold the carrier itself together, it will just carry the container hold by the ropes connected to the winches on the top of the carrier. The point is that the force caused by the weight of the bridge and container is compressing the top and therefor bending the pillars - you can compare it to a beam bridge- would you like to drive such a bridge with a car when the pillars are bending? The yellow one is forwarding to force to the ground better since the track is also directly in the center of the pillars. 

You need to make the top and the pillars much more rigid, your grey liftarms are a first step on a longer way.

See a real life description of such a beam bridge: https://science.howstuffworks.com/engineering/civil/bridge3.htm

A nice challenge - stay on it and good luck!

 

Oh, this model is very far from being accurate to the real thing. Having the wheels be correct would be nice, but I have no hope of getting a string-based lifting system to work.

Here in January I unfortunately don't have access to my collection. I am, however, so satisfied with how the spreader works, that I have made building instructions for it:

https://brickhub.org/i/709

It can be made prettier with end caps: https://brickhub.org/i/708

but as explained in this video, end caps make it unusable when trying to stack containers back-to-back: