[CONCEPT] All wheel steering and crab steer as well

[Leif]

As so many of my builds this is a crude mockup of an idea that I have been pondering about and had to get out of my head, down to the table!

This steering setup supports 4 axle steering and crab steer as well.

Naturally Ackermann went out the window but I think we can make it at least this time

What I have done is that the feed is from the first or the second motor, to differentials - backwards use of differentials so to speak - and output from differentials will support rack and pinion steering.

Note that the two middle diffs are geared down in order for proper radius thinking when turning, but all diffs turn at the same speed and direction when crabbing.

8x8x8 will set you back 8 differentials but that's another story

[Alasdair Ryan]

It's a interesting concept,but I think the wheels will go out of sync if they hit a object.

[Leif]

It's a interesting concept,but I think the wheels will go out of sync if they hit a object.

Thank you for your kind words. You may be correct but I cannot see that issue as it is now. The motor locations are temporary and they should of course drive via worm gears in next version.

Edited October 19, 2014 by Leif

[dhc6twinotter]

I drew up some sketches for a similar setup several years ago, but I tossed the idea because of some issues with the design. I never got around to building a prototype, and I hope this works well for you. It's an interesting concept.

The issues I found:

-All the gearing meant there was slop in the system and wheels may wobble quite a bit.

-Steering had to return to center perfectly before switching modes, otherwise the wheels would lose sync with each other.

But again, I never actually built a model to see if my concerns were justified. I'm curious to see if this works for you, and I hope it does. It's a neat idea.

[Alasdair Ryan]

The issues I found:

-All the gearing meant there was slop in the system and wheels may wobble quite a bit.

-Steering had to return to center perfectly before switching modes, otherwise the wheels would lose sync with each other.

I think you could get over both of those issues with using worm gears before each axle,but you could only do this per one side.

Two pneumatic rams per each axle and a center return valve switching crab and all wheel steer may be a better route to go down.

[Lipko]

I ditched the idea of solving the crab/4-wheel steering mechanically because of the slack in the system. But some time ago someone posted an awesome idea to solve it: Use two pieces of steering servo motors for each wheel-set, and simply add a pole reverser to the motor that's moving the rear axles (or just use the reverse button on your remote control unit).

By this way you have two independent motors that you can control at the same time (by connecting the remove control arms for example), you can switch steering mode even if the axles are steered, the electronics will sort everything out.

The only reason I don't use this system is that I'm building a manual model.

Edited October 19, 2014 by Lipko

[Leif]

I ditched the idea of solving the crab/4-wheel steering mechanically because of the slack in the system. But some time ago someone posted an awesome idea to solve it: Use two pieces of steering servo motors for each wheel-set, and simply add a pole reverser to the motor that's moving the rear axles (or just use the reverse button on your remote control unit).

By this way you have two independent motors that you can control at the same time (by connecting the remove control arms for example), you can switch steering mode even if the axles are steered, the electronics will sort everything out.

The only reason I don't use this system is that I'm building a manual model.

I am not sure I understand completely.

4 axles, 1 servo motor for the first 2 axles, and 1 servo for remaining 2 axles? Or another setup?

Remember, during normal steering the 2nd and 3rd axle do not turn as much as 1st and 4th, but they all turn the same length when crabbing.

Generally speaking, servos will of course help to auto-center, before switching steering mode. That is a question of taste; jerky steering with servos (or train remote), or smooth steering with medium motors...

[Zerobricks]

You can solve sync problem by using servos Than it all comes down to control

[Rockbrick]

by trying on the table and seeing things not work straight away is how we learn better to solve the problem,

looking good so far...

I converted my 8865 once to 4WS - and I had a gearbox on the steering to select full opposite lock steering, honda prelude style 4WS (rears turn a little bit in relation to the front) and full crab steering but It only worked limiting it a bit as the ackerman would take over and mess it all up - you had to change the selector when all wheels were in a straight line too

Edited October 20, 2014 by Rockbrick

[TJoling]

I've been tinkering with this idea before too. I built an adder-subtractor construction inbetween the axles, and used both outputs to steer the front and the rear axles separately. In essence it's the same system you built, only now you save yourself two diffs. But again, as said before, I also discarded the idea due to slack in the system and having to perfectly align into the straight position before switching to the other mode. I made a small demo back then.

I have also been thinking about the ackermann in normal steering and non ackermann in crabsteering and how to solve that. at some point, Barman came up with a great idea:

But to mechanically control this is terribly complicated, you're almost forced to use mindstorms.

[Leif]

I've been tinkering with this idea before too. I built an adder-subtractor construction inbetween the axles, and used both outputs to steer the front and the rear axles separately. In essence it's the same system you built, only now you save yourself two diffs. But again, as said before, I also discarded the idea due to slack in the system and having to perfectly align into the straight position before switching to the other mode. I made a small demo back then.

I have also been thinking about the ackermann in normal steering and non ackermann in crabsteering and how to solve that. at some point, Barman came up with a great idea:

But to mechanically control this is terribly complicated, you're almost forced to use mindstorms.

You have some good ideas there but I wonder, with your 2diff system, do the pair of axles turn the same degrees?

In crab all axles turn the same degree, but normal steering, inner axles (2 and 3) should turn less than outer axles (1 and 4) in order to get as little tyre rubbing as possible, even without Ackermann.

My system with 4 diff takes this into consideration. perhaps you didn't see it from my pic but I gear down for axle 2 and 3 and gear up again for axle 4, for normal steering..!

see this pic, two options:

[TJoling]

I did notice the change in steeringangles between the modes in your system. However, to achieve that you have to make the system a lot bigger, and the slack will even further increase as you're using a lot more gears. I tried to solve said problem by putting a 12z gear on the gearrack of the first axle seen from the middle, and a 14z gear on the gearrack of the second axle seen from the middle. This reduced the slack between the axles in a pair to a minimum, and made the axles approach the right angle in either mode.

I think that it depends on the size and weight of the vehicle what would be the best solution in this case.

[Leif]

Alright, got it, but how can you have identical angles on four axles in crab mode if you have different gear on different rack?

I browsed through your flicker, and that's a good machine you building there!

do you have any more detailed pics about the steering?

Oh and btw, yeah slacking is always an issues with the gears we have..as few as possible and as few perpendicular as possible as well, that's what we have to aim for!

My concept build in this thread can of course be optimized at a few places and the crab gears do not really slack, it's the normal turn side that slacks, but I have some ideas. perhaps divide with two engines

and of course no gearing up to fourth axle, that was just a lazy and quick thing!

Edited October 20, 2014 by Leif

[TJoling]

Thank you, I've been working on it for quite a while now, but the end isn't anything near in sight

Luckily, I have the original middlesection with the adder-subtractor assembly still in one piece. I will try to make some detailed pictures of it when I'm home this weekend.

To answer your first question: I said the axles approach the right angle in either one of the modes The ratio between the axles is 6:7, so pretty close to 1. This means the axles don't have the right angle in normal steering, or in crabsteering, but both ways the angles are pretty close. The slack between the gearrack and wheels, together with the all wheel drive, will give the wheels enough play to reach the optimum angle on their own. In theory, it's more like crapsteering , but in practice it works pretty fine.

Your mock-up could indeed be improved in terms of slack, but I still think the biggest causes of slack are the diffs, which can not be discarded in your design. I would still like to see an improved version though

Edited October 20, 2014 by TJoling

[Leif]

lol at "crapsteering"

Yeah I got another machine 8x8x8 with PGA - Pretty Good Ackermann...

Slack is always an issue with our gears, I am pondering on using worm gear after each diff, that would remove the slack for the rack and pinion part.

Issue in that case is of course that I alter the gear train by 90 degrees so I have to build a new setup in order to get to the gearrack, so the hits keep on coming..

[That_LEGO_Guy]

I was sure I'd seen something with Crab/Ackerman before and finally found it, it's part 6 below the picture. Link to topic http://www.eurobrick...pic=89963&st=25

Hi Orbit,

Sorry I missed your reply the first time. Here are the answers I have:

1) I haven't measured the weight, sorry. But it is heavy... I usually transport it in three pieces: the undercarriage, the superstructure and the boom. Each section is several kg's...

2) The outriggers are not strong enough to lift the model. When fully extended they do bend. They do however improve the stability during rotation of the superstructure.

3) I still have to finish the boom, but it will get at least 4 sections. The first section is the outer shell, the second section is actuation with a long gear rack. Section 3 and 4 are actuated with strings, just like the 42009 boom. Each section is ~80 cm, so the total height will be around 3 meters.

4) In the pictures posted the actual shifting mechanism wasn't visible. Here is a view from the bottomside of the switching mechanism. The gears are switched in groups of 3. The black knobwheel in the middle is elastically locked by the red bionicle teeth. It can be positioned in 4 different positions. For each position a set of 3 gears is engaged.

5) I haven't completely solved that yet. Reinforcing the drivetrain has helped a lot, but not enough. Adding elastic energy through elastic bands was my next trick. That helped, but it put a big strain on the structure so I wasn't too happy about that.

6) For the normal steering I use the same mechanism as the 42009 set: each gear rack is at a different distance from the wheel axle. One long steering axle can be turned to steer the wheels proportionally to their position.

When switching from normal steering to crabsteering the whole steering axle shifts 2 studs to the back. All the 12T gears disengage from the normal steering racks. Other 12T gears on that same axle engage to the second set of gear racks. This second set is located two studs behind the wheel axle for each wheel.

When turning the steering axle now all wheels will turn the same direction.

I hope this answers your questions.

Edited October 22, 2014 by That_LEGO_Guy

[Leif]

I was sure I'd seen something with Crab/Ackerman before and finally found it, it's part 6 below the picture. Link to topic http://www.eurobrick...pic=89963&st=25

Holy Moly that was a big machine! did he ever finish the beast?

A really interesting concept with double gear racks for steering, I will steal with pride

Don't wanna bash the guy, and he confirmed in the last post as well, there is no Ackermann in this setup.

The main thing that has me perplexed is if it's true Ackerman steering geometry or not. Upon first inspection one might say "look at the pivot point at the wheel and the pivot point where the gear rack is connected", but upon further analysis those do not appear to be the correct pivot points. Since the 13L gear rack makes a rectangle (a rhombus when steered), I feel that the two beams leading to the 7L gear rack also make a rectangle / rhombus, not a trapezoid as is required for Ackerman steering geometry.

No you are not wrong. The right and left wheels do have the same steering angles. The angles per axle are different though (when not in crabsteering mode), so it is more like an Ackermann-light :).

But that's alright, we cant win them all! not without hydraulics...

[Jeroen Ottens]

Holy Moly that was a big machine! did he ever finish the beast?

A really interesting concept with double gear racks for steering, I will steal with pride

Don't wanna bash the guy, and he confirmed in the last post as well, there is no Ackermann in this setup.

It is still in progress (I actually solved the problems I had with the outriggers and I have found a neat way to add some elastic energy to the boom to help during lift), but all these contests here take up a lot of time

I wouldn't say no Ackerman, but it is Ackerman-light, since the wheels do turn at different angles going from the front to the back. In practice the turning angle is so big that the different angles for the left and right wheels (like you would need for a true Ackerman steering) can be found within the tolerances of the Lego.

But feel free to steal, that's what this forum is for

[Leif]

It is still in progress (I actually solved the problems I had with the outriggers and I have found a neat way to add some elastic energy to the boom to help during lift), but all these contests here take up a lot of time

I wouldn't say no Ackerman, but it is Ackerman-light, since the wheels do turn at different angles going from the front to the back. In practice the turning angle is so big that the different angles for the left and right wheels (like you would need for a true Ackerman steering) can be found within the tolerances of the Lego.

But feel free to steal, that's what this forum is for

Good to hear about progress!

Yeah I think I have been studying too hard on steering geometry for each and every automobile so I forget that it really doesn't matter all the time to pinpoint all those darn angles

I built a big beast superstructure 8x8x8 using Unimog tires and on this one I actually did Ackermann as well, and what do you know - I can drive and turn on a thick turkish carpet, even with the superstructure and big boom on top! so little skidding, next to noe

http://www.eurobrick...opic=100552&hl=

It is satisfying to see all those calculations pay off from time to time!

But current build will be more or less exactly like yours, different steering angles for every axle, normal + crab (If I solve all issues) and we go for "Ackermann light". I will also use the 42009 wheels + more so skidding will not be such a big issue as the Unimog tires are, especially with portal hubs as well..

Now down to the man cave and look into new crab solution!

keep us posted!

[Leif]

@Jeroen, do you have some info or LDD about the mechanical block you have to prevent steer mode switching when not in straight mode?

I would really appreciate some insight there!

[Jeroen Ottens]

@Jeroen, do you have some info or LDD about the mechanical block you have to prevent steer mode switching when not in straight mode?

I would really appreciate some insight there!

Here it is:

The red parts are part of the steering axle (which can be pushed two studs when switching sterring mode). The yellow bushes have to travel in between the two 5x3 L beams. Even when the axle is not perfectly aligned it will straighten itself when switching modes.