Steering mechanism for tight wheel wells?

[AussieJimbo]

Couldn't help myself, kept playing and managed to get a rough proof of concept working based on the above.

Now very, very AM here so photos will have to wait until tomorrow.

:classic:

[fmmjqtft]

Here is my design. Each pivot axis remains in the exact centre of its wheel, but the wheels move inwards slightly as the vehicle steers.

Assumption to be made:

1. The blue friction pin never slips.

2. The position of the blue friction pin is fixed relative to the vehicle, but it can freely rotate about its axis of symmetry.

3. The position of the central '3L Axle with stop' is fixed, but it can freely rotate about its axis of symmetry.

The reason for using a friction pin here is that the 16t gear has to be rotated through 11.25 degrees, which is difficult to do with other parts. The wheels are both offset by 0.5L so that the axis of symmetry is central.

Top:

Bottom:

I will upload an ldr file and draw a diagram later.

Edit:

Here is the ldr file:

--------------------------------------------------

0 Untitled

0 Name: steering1.ldr

0 Author: MLCad

0 Unofficial Model

0 ROTATION CENTER 0 0 0 1 "Custom"

0 ROTATION CONFIG 0 0

1 0 0 -50 30 0 0 1 0 1 0 -1 0 0 32316.dat

1 0 0 -50 -10 0 0 1 0 1 0 -1 0 0 32316.dat

1 0 0 -50 -50 0 0 1 0 1 0 -1 0 0 32316.dat

1 14 -80 -30 -10 0 0 1 0 1 0 -1 0 0 55982.dat

1 14 80 -30 -10 0 0 -1 0 1 0 1 0 0 55982.dat

1 7 -40 -30 10 -1 0 0 0 0 -1 0 -1 0 48989.dat

1 7 40 -30 10 -1 0 0 0 0 -1 0 -1 0 48989.dat

1 7 0 -70 -10 -1 0 0 0 0 -1 0 -1 0 4019.dat

1 7 0 -70 -50 -0.980785 -0.19509 0 0 0 -1 0.19509 -0.980785 0 4019.dat

1 8 0 -30 -10 -1 0 0 0 -1 0 0 0 1 33299.dat

1 28 0 -50 -10 0 -1 0 1 0 0 0 0 1 6587.dat

1 28 60 -30 -10 -1 0 0 0 0 -1 0 -1 0 6587.dat

1 28 -60 -30 -10 1 0 0 0 0 -1 0 1 0 6587.dat

1 1 0 -60 -50 0 -0.980785 -0.19509 -1 0 0 0 0.19509 -0.980785 3749.dat

1 0 40.275 -10 -29.74 -1 0 0 0 1 0 0 0 -1 32523.dat

1 0 -39.725 -10 -29.74 -1 0 0 0 1 0 0 0 -1 32523.dat

1 0 80 -30 -10 0 0 -1 0 1 0 1 0 0 30391.dat

1 0 -80 -30 -10 0 0 -1 0 1 0 1 0 0 30391.dat

1 19 40 -30 -50 0 0 -1 -1 0 0 0 1 0 32556.dat

1 19 -40 -30 -50 0 0 -1 -1 0 0 0 1 0 32556.dat

0

--------------------------------------------------

Please note that this particular design will not work if built, but it should be easy to modify.

Edited June 25, 2011 by fmmjqtft

[AussieJimbo]

Alright, here is the 6-wide proof of concept. Please ignore the superfluous sticky out bits (2x8 black plate, 7 wide black beam, big overhang at the front), I was scratching about for parts.

At the moment the steering is reversed (turn left, goes right) but that's trivial to fix with some gearing. The grey bricks stop the long push rods (dark grey beams) from moving outwards too far and indicate where the sides will be in a future model.

I needed some jumper plates (in white, hard to see) to shift to vertical steering axis half a stud for correct alignment.

^Large

^Large

^Large

A side effect of the design is that at full lock, the wheel being pulled (the higher wheel in the shot below) gets overdriven (turned too far). The effect of this is to decrease the turning circle so I see it as a good problem to have (at the cost of slight scuffing) but of course you could prevent it by simply limiting the available lock.

^Large

Look out for a steerable 6-wide in the Town forum some time soon.

:classic:

Edited June 26, 2011 by AussieJimbo

[fmmjqtft]

Here is an animation:

Here is the diagram:

All blue angles have the same value. The blue and red line segments have the same length as the circle's diameter.

[Sokratesz]

Here is an animation:

Here is the diagram:

All blue angles have the same value. The blue and red line segments have the same length as the circle's diameter.

I really, really like that. Any chance it can be made smaller than 5x5? (without the wheels)

Edited June 26, 2011 by Superkalle
Removed pics from quote

[fmmjqtft]

I really, really like that. Any chance it can be made smaller than 5x5? (without the wheels)

It probably is possible to make a working model smaller than 5x5LL, but it would have to be taller. Note that this particular construction will not work if built (the blue axle pin will slip), but it can probably be modified to work with an extra 1L of height.

To make a smaller version, it may be easier to turn the red and green parts from further away, using two axles, instead of trying to build two gears onto the construction.

[AussieJimbo]

That looks like a good solution. To avoid the blue pin slipping, you could replace it with a 3 axle and lock it underneath with a 2 x-hole half thickness beam with the blue pin in the next hole along.

:classic:

Edited June 26, 2011 by AussieJimbo

[fmmjqtft]

That looks like a good solution. To avoid the blue pin slipping, you could replace it with a 3 axle and lock it underneath with a 2 x-hole half thickness beam with the blue pin in the next hole along.

:classic:

The 16t gear needs to be rotated through 11.25 degrees relative to the 5L beam, which is why I used an axle pin. An easy solution would be to lock an extra 16t gear (using a 2x1x0.5 beam as you said), next to this gear, which would keep this gear rotated at 11.25 degrees.

[Sokratesz]

And then all this needs is a rubberband construction so it RTC's, and you can build a beautiful small RC car :)

[Superkalle]

This discussion is turning out really, really well. Thanks all for you clever input. Specially like you drawing and animation fmmjqtft.

Now, a side question to you all - is it really so that this topic has not been discussed anytime earlier, on EB or elsewhere? It just doesn't seem possible that this is the case.

[fmmjqtft]

Here is a working version.

[DLuders]

Here is my LDD steering solution for a 5-studs-wide Lego car. The wheels pivot on the wheel hubs' true centers by rotating the central gear, but I admit that the assembly is a bit tall. Perhaps somebody knows how to make a lower version. Here is my LDD .lxf file:

Mini_Steering_3_DLuders.zip

[DLuders]

I wrote an e-mail to Lego911 (a master Lego car builder who works for Ford Australia), and got this reply:

"I looked at the Superkalle steering challenge link. It looks like they are almost there.

"I have a relatively good solution. More importantly, it solves the question people are asking - 'How do you make a steering mechanism that does not drive the wheel fore and aft under steering lock?'.

"There is an implication that 'real' cars do not do this - this is incorrect, most do. They can play with the king pin inclination angle by moving the upper pivot inwards. Though this creates the point of rotation of the tyre on the ground, in the centre of the tyre face, it still allow the entire mechanism to more backwards and forwards. For very wide-wheeled cars, they attempt to mount the pivot points within the wheel volume, as close to the hub centre as possible, but very rarely is it in the centre axis of the tyre.

"BMW, Audi, and Ford Australia have a system that can achieve this. It is called a 'double, virtual pivot axis'. It achieves the effect by having two lower suspension links, at an angle to one another, and with their outer end pivot points offset from one another (as viewed from above). This creates a virtual pivot axis at the wheel end (on the lower link plane). The angle of the arms, along with the linear displacement of the ends, defines the 'virtual pivot point'. That is, in the plane of the lower arms, the virtual pivot point is the point described by the point of intersection of the two arm axes. This point will actually move through the suspension movement (not by much) and will actually drive the hub forward and backwards. The tyre contact patch will be centered on the rotation axis of the hub.

"This particular steering arm type was selected due to the steering performance requirement, along with a need to package front drive shafts for the four-wheel-drive installation in the CUV (Ford Territory) model. The drive axle mechanism prevented the fitment of arms close within the wheel volume.

"The LDD Lego model attached illustrates the mechanism. Instead of two lower pivot arms, I transposed the second arm to provide the top link. As seen from above, you can see the angle positions and the linear (hub) offset. The virtual pivot is created by the intersection of the arms as seen from above. I have copied the mechanism, and steered the wheel hub. You can see that now, the arm's angles have changes, and that the virtual pivot point has also moved forwards (or backwards). However, the centre of the tyre is still in the same position (fore-aft). Although there is not a massive amount of lock available, the wheel still stays centered in the wheel-well. The mechanism is strong enough to support the vehicle weight of the cars that I design - though it is not very play-robust.

"As a comparison, I have created a similar dimensioned, single pivot axis assembly for comparison. The lock angle is similar, but the wheel/tyre volume moves a long way forward and back, into the body work.

"Although my design does not meet the criteria of steering 'control', and at this scale, does not fit in an 8-wide (I can make an 8-wide systems that compromises the geometric purity of this model) - it does answer one of the discussed fundamental questions of real auto-design steering mechanism geometry.

"Feel free to pass on to the Eurobricks form if you like. For builders who want the steering effect but not the 'playability', this mechanism might be exactly what they are after."

Double_Pivot_Steering_Lego911.zip

[Superkalle]

Here is fmmjqtft's solution as a LXF. (Hope you don't mind fmmjqtft?)

I tried it out with physical bricks. It works well, but naturally with all the joints and cogwheels it does have a fair amount of play. I'm testing if and how it can be incorporated in a vehicle.

PS: I now understand what you mean by 11.5 degrees. Clever way you solved that. Question though - what would happen if you skipped that compensation, as it adds much to the complexity of the solutio (i.e. what would be the consequence?)

SteeringByfmmjqtft.zip

[AussieJimbo]

Thanks for that post from your bloke at Ford, DLuders. Very informative. Your design looks interesting.

fmmjqtft, that refinement at the top of the page looks very good.

Here is the latest version I've been working on. It meets the rotation requirement with narrow wheels, with the wide ones it's a bit out but the movement is still well contained.

I simplified the steering linkage and tidied up the support structure. This iteration has proven smooth and reliable, standing up well to multipoint turns and enthusiastic driving around the coffee table.

With the narrow wheels it could be hidden in a 6-wide with a bit of tile work around the front but I think I'm going to leave the wide wheels on it and build a chunky 8-wide truck of some kind first.

^Large

^Large

^Large

^Large

:classic:

Edited June 27, 2011 by AussieJimbo

[fmmjqtft]

Here is fmmjqtft's solution as a LXF. (Hope you don't mind fmmjqtft?)

I tried it out with physical bricks. It works well, but naturally with all the joints and cogwheels it does have a fair amount of play. I'm testing if and how it can be incorporated in a vehicle.

PS: I now understand what you mean by 11.5 degrees. Clever way you solved that. Question though - what would happen if you skipped that compensation, as it adds much to the complexity of the solutio (i.e. what would be the consequence?)

The rotation of 11.25 degrees is required. I could do an animation showing what it would be like without it.

I will upload a more compact version. My second version of this design was modified so that larger wheels could be used, so that it could fit into a vehicle like this:

[fmmjqtft]

Here: A similar modification can also be made to my other design (at the top of this page).

Here is the ldr:

----------------------------------------

0 Untitled

0 Name: steering2.ldr

0 Author: MLCad

0 Unofficial Model

0 ROTATION CENTER 0 0 0 1 "Custom"

0 ROTATION CONFIG 0 0

1 7 -40 -30 10 -1 0 0 0 0 -1 0 -1 0 48989.dat

1 7 40 -30 10 -1 0 0 0 0 -1 0 -1 0 48989.dat

1 28 60 -30 -10 -1 0 0 0 0 -1 0 -1 0 6587.dat

1 28 -60 -30 -10 1 0 0 0 0 -1 0 1 0 6587.dat

1 7 -55 -30 -10 0 0 -1 -1 0 0 0 1 0 32123a.dat

1 7 55 -30 -10 0 0 -1 -1 0 0 0 1 0 32123a.dat

1 14 0 -45 -10 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 14 0 -55 -10 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 14 0 -45 30 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 14 0 -55 30 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 14 0 -55 -50 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 14 0 -45 -50 0 0 -1 0 -1 0 -1 0 0 32017.dat

1 0 0 -35 30 -1 0 0 0 1 0 0 0 -1 6632.dat

1 0 0 -25 30 -1 0 0 0 1 0 0 0 -1 6632.dat

1 14 -40 -10 -30 -1 0 0 0 1 0 0 0 -1 32523.dat

1 14 40 -10 -30 -1 0 0 0 1 0 0 0 -1 32523.dat

1 14 0 -90 -30 -1 0 0 0 1 0 0 0 -1 32523.dat

1 8 0 -70 -10 -0.980785 -0.19509 0 0 0 1 -0.19509 0.980785 0 6542.dat

1 7 0 -70 30 -1 0 0 0 0 1 0 1 0 4019.dat

1 7 0 -70 -50 -1 0 0 0 0 1 0 1 0 4019.dat

1 7 0 -50 30 0 0 1 1 0 0 0 1 0 4519.dat

1 0 0 -60 -10 0 0 1 1 0 0 0 1 0 3706.dat

1 0 0 -60 -50 0 0 1 1 0 0 0 1 0 3706.dat

1 7 40 -30 -50 0 0 1 1 0 0 0 1 0 87082.dat

1 7 -40 -30 -50 0 0 1 1 0 0 0 1 0 87082.dat

1 8 0 -30 -50 0 0 -1 0 -1 0 -1 0 0 60483.dat

1 0 -20 -40 -50 0 0 -1 1 0 0 0 -1 0 3673.dat

0

----------------------------------------

Edited June 27, 2011 by fmmjqtft

[fmmjqtft]

I tried it out with physical bricks. It works well, but naturally with all the joints and cogwheels it does have a fair amount of play. I'm testing if and how it can be incorporated in a vehicle.

So far my design fails . Today I tried building it into a vehicle similar to 8635, but it only worked when suspended in air. The construction bends too much when the wheels touch the ground.

It is difficult to build a working model at such a small scale. Anyone who wants to give it a try can use this diagram:

It is a parallelogram where two sides must be rotated in the opposite direction to the other two sides, relative to the vehicle. The blue angles have the same value.

[Richard L aka Scooby]

Dear all,

This is my version and it works perfect.

Edited June 30, 2011 by scooby136

[Mephistofeles]

Dear all,

This is my version and it works perfect.

Yeah, perfect, so why do you have such a large wheel arch?

[garson]

have you tried this http://technicbricks.blogspot.com/2011/04/week-techvideo-2011-15-city-hunter.html

[Richard L aka Scooby]

Yeah, perfect, so why do you have such a large wheel arch?

There is no large wheel arch. When the steering is straight forward then the wheels fitting precise in the wheel arches. Build the same model what i have made then you wil see. Then judge.

[Silcantar]

Peter Blackert's (lego911's) design really caught my interest, so I decided to try it out myself. Although it won't work for Superkalle's MOC, it works quite elegantly at larger scales. The picture is crappy (I took it with my cell phone using ambient indoor light >_<), but it gives you an idea of what I did:

The design is really very effective. In this case, it is actually possible to make the wheel arches smaller than the maximum diameter of the balloon tires.

I also made a prototype of the system using multilink independent suspension.

I also experimented some with camber and caster on this one. As a result, the inside wheel when turning of a car with this suspension geometry would have positive camber, while the outside wheel would have negative camber. I am looking forward to possibly implementing this suspension geometry on one of my MOCs in the near future.

[Superkalle]

So far my design fails . Today I tried building it into a vehicle similar to 8635, but it only worked when suspended in air. The construction bends too much when the wheels touch the ground.

You're right - I only tested it in mid-air. Thanks for posting.

There is no large wheel arch. When the steering is straight forward then the wheels fitting precise in the wheel arches. Build the same model what i have made then you wil see. Then judge.

Mephistophelses was a bit though on you, but in essence he is right. The challenge in this topic is to create a steering mechanism for tight arches (aka wheel well), meaning that in your case, the arch should not be no more then 4 studs. And you shouldn't be able to see the steering mechanism on the side the way you can on your truck.

Anyway, nice Scania

I am looking forward to possibly implementing this suspension geometry on one of my MOCs in the near future.

If you do, please make a quick post in the topic as well, so we can see how it turned out

[le60head]

Looking forward to seeing that!

Hello!

First of all, i wanted to thank you all for the information you shared in this topic!

It served as a great source of inspiration to me.

After months of delay, caused by me being lazy, i finished the model i was working on.

The steering system i used doesn't quite fit your description. The wheel wells on my model are 5 studs after all, which is huge.

Yet it is the most compact motorized steering system i've done so far. The wight of the entire model is just 7 studs.

I've made a video in order to explain how it works.

Please look at my topic to see it used in my model of a Blue Bird School bus.

LEGO BLUE BIRD All American School Bus

Regards,

legohead

Edited November 23, 2011 by le60head