ETF HTX-380 Mining Truck.

[Moz]

@ too expensive ... every four PF-L you need (in best case) an receiver and a battery box.

Ah, I assumed that like a lot of Technic builders you have a stack of those lying around. I have the 6-AA battery boxes in silly numbers from my road train project, since I bought several 8043 (excavator) and 8258 (crane truck) for the M motors and other parts. When I tried to sell some battery boxes off no-one was interested. I ended up buying extension leads as spare parts, that was unexpectedly expensive. Oh, and 2x2 round bricks to reduce wind-up in the axles (they make a significant difference).

@ackermann ... the difference should be enough to get a proper ackermann geometry at the second axle. the insane is, to find out if it is correct i have to build the whole chassis.

?? I'm not sure why, I just built up a frame approximately the right size and mounted a single wheel under each turntable. Putting that on sheets of paper on the floor and rolling the model across it showed an awful lot of scrub no matter what I did. The "paper on the floor" trick is what I use for recumbent trikes, BTW. But that does mean you need the turntables and much of the steering mechanism worked out. Once I gave up on simple levers and started using gearing it actually got easier, because I could shift to using worm gears driving the turntables and wind-up in the axles didn't matter as much. But it meant changing gears between turning left and right, which is where the cams came in (make sure the change happened exactly straight). For a while I thought I could get crabbing by using 5 gearboxes, which I suspect is actually possible, but that was about the time I gave up. I was using simple geometric ratios, but they aren't constant over the steering angle so it all got a bit hard. I kept imagining some funky mechanical integrator arrangement that could get it right, but I'm not smart enough to work it out - I grew up with digital controllers, so I'd use maths directly.

Here's what I mean by geometric - I sketched the front half of the vehicle when it was turning at twice its width (because that's easy to visualise). The grey lines intersect at the centre of the turn. I used this as my steering limit, BTW, and drew another couple of sketches at intermediate angles to see how bad the error was if I used simple, constant ratios.

[efferman]

yes, i have roundabout 15 batteryboxes and many many PF-M. that is absolutely too much, because i dont want use so many of them.

when i look at your sketch with the turning circle, it looks that i will have big differencies in wheelspeed between the multiple axles. so i have to make good decission by planning the differentials to get no rubbing wheels. again i see the need to build the whole chassis in real because i want that it works good enough for other people.

[PsyKater]

What a great truck.

ETF developed a unique steering system. All wheels are steered at low to medium speeds reducing tyre wear, while at higher speeds the last two axle lines gradually change to nearly rigid resulting in better stability. ETF trucks are equipped with a special speed-proportional steering system. At very low speeds, such as when spotting the loader, the system requires only two turns from lock to lock. The whole system works automatically without interference from the operator..

http://ich.tvbvideo.de/video/ab270399248s.html

I like this project a lot and I'd love to see this thing in LEGO (maybe even before it exists for real). But I fear it is too expensive...

[efferman]

i have a little problem, can me someone explain how i can retract the linear actuators to fit them in their holes?

[Alasdair Ryan]

I can do that for you if you can post the LDD file.

[efferman]

thank you very much for your service, but i want be able to do it myself. but here it is

in meanwhile i have found something, what is working acceptable. the only problem is that my LAs should be between 2,25 and 2,50, so i must use axles instead of pins.

Edited January 30, 2013 by efferman

[Alasdair Ryan]

Ok in that case download the different La lengths here.

[efferman]

this is what i have used

[Alasdair Ryan]

Then you should just have to angle and pick the right size.

[efferman]

honorable scotsman, thank you very much for your advice, but please read

..... the only problem is that my LAs should be between 2,25 and 2,50, so i must use axles instead of pins.

[Alasdair Ryan]

I don't understand what the problem is.....

Edit is it the bottom part of the La hitting the lift arm frame?

If so you could use a smaller one.

Edited January 30, 2013 by Alasdair Ryan

[D3K]

Holy, I am trying to wrap my head around MATLAB as we speak! Hope to never have to resort to that tool for building Legos:-)

EDIT;

A little late with that post I see...:-P

Edited January 30, 2013 by D3K

[Alasdair Ryan]

Holy, I am trying to wrap my head around MATLAB as we speak! Hope to never have to resort to that tool for building Legos:-)

EDIT;

A little late with that post I see...:-P

Yeb

[efferman]

I don't understand what the problem is.....

Edit is it the bottom part of the La hitting the lift arm frame?

If so you could use a smaller one.

no sir, in real life it works properly, only the limited functionality of ldd is a problem

[Alasdair Ryan]

Could you not just send me the file?

[efferman]

please follow the orange markers

thank you very much for your service, but i want be able to do it myself. but here it is <<<<

in meanwhile i have found something, what is working acceptable. the only problem is that my LAs should be between 2,25 and 2,50, so i must use axles instead of pins.

[Alasdair Ryan]

please follow the orange markers

How did I miss that,ok What is the problem can you circle around the problem?

[efferman]

in the ldd file which you have in this moment, i have no problems anymore. only the green axle is a replacement of a 3l pin which dont want fit.

btw, iam actual searching the correct linkage to get a proper ackerman geometry.

p.s. its not final

[Saberwing40k]

How did you all get those LA's in different lengths in the first place? That trick would be useful to know.

Anyway, no pictures yet, but I am making a new axle, only sort of like efferman's. This has turned into an interesting rivalry.....

[efferman]

How did you all get those LA's in different lengths in the first place? That trick would be useful to know.

i have found them here

http://www.eurobricks.com/forum/index.php?showtopic=52005entry925927

[Moz]

when i look at your sketch with the turning circle, it looks that i will have big differencies in wheelspeed between the multiple axles

About 3:1 in the above diagram, but that's why you use one electric motor per wheel pair. The slowest wheels will almost stall, while the fastest will not provide much push. If you try to use fewer motors you will need a lot of differentials. The intermediate gearing between the diffs will produce a sort of limited slip effect which will be helpful.

If you use a LA to change the leverage of the main steering LAs you should be able to halve the slip error without much work, but the mechanism I found was bulky. Getting the right part of the sine curves just with single LAs and levers was tricky, IIRC the proper lever length is about 3 studs for a 45° each way turning wheel so slop becomes a big issue.

Using gears driving the turntables directly simplifies problems with extreme steering angles, but I suspect you'd want to use something like MatLab to plot the 4 steering angles vs turn radius then try to get the best straight line fit to each curve. Using gearboxes, I built one output per line then swapped them between wheelsets depending which way I was turning (ie, the left and right wheelsets swap over depending on whether they're the inside wheelset or the outside one). This could all be done entirely mathematically, and I suspect some smart person could actually derive it from first principles into a clean set of equations. Then all you need is three continuously variable, no-slip gearboxes :)

Edited January 31, 2013 by Moz

[efferman]

About 3:1 in the above diagram, but that's why you use one electric motor per wheel pair. The slowest wheels will almost stall, while the fastest will not provide much push. If you try to use fewer motors you will need a lot of differentials. The intermediate gearing between the diffs will produce a sort of limited slip effect which will be helpful.

why not, this huge thing has enough room for a complex drivetrain. and if it doesnt work, there is enough room to give each wheel its own PF-L.

If you use a LA to change the leverage of the main steering LAs you should be able to halve the slip error without much work, but the mechanism I found was bulky. Getting the right part of the sine curves just with single LAs and levers was tricky, IIRC the proper lever length is about 3 studs for a 45° each way turning wheel so slop becomes a big issue.

Using gears driving the turntables directly simplifies problems with extreme steering angles, but I suspect you'd want to use something like MatLab to plot the 4 steering angles vs turn radius then try to get the best straight line fit to each curve. Using gearboxes, I built one output per line then swapped them between wheelsets depending which way I was turning (ie, the left and right wheelsets swap over depending on whether they're the inside wheelset or the outside one). This could all be done entirely mathematically, and I suspect some smart person could actually derive it from first principles into a clean set of equations. Then all you need is three continuously variable, no-slip gearboxes :)

thanks for this exciting excursion in the world of maths, but iam not an theoretical. there is a simple way to get a proper ackerman geometry.

choose a point on the length line of your vehicle. then use some beams which are long enough to reach the pivot points of your steering wheels, and fix it to both points. on this line should u choose the point for connecting the track rod. if this is done properly the ackermann works proper. you only need to have the chassis.

[Moz]

there is a simple way to get a proper ackerman geometry.

choose a point on the length line of your vehicle. then use some beams which are long enough to reach the pivot points of your steering wheels, and fix it to both points. on this line should u choose the point for connecting the track rod. if this is done properly the ackermann works proper. you only need to have the chassis.

I assume you mean something like this: https://en.wikipedia.org/wiki/Ackermann_steering_geometry

That works, especially since Lego models (and to a large extent the real truck) fit the "low speed" requirement. We also don't really care about suspension travel.

The problem is steering angles/minimum turn radius. Ackermann limits this quite a lot, with flop at travel limits being a secondary concern. I think you'll struggle to get to the limit in my diagram above.

[efferman]

indeed, roundabout at this angle will be the maximum

100_9962 von efferman auf Flickr

Edited January 31, 2013 by efferman

[Feradin]

@ efferman

I'm really impressed. Would replicate the model as quite interesting.