Pushing a Locked / Unlocked Telescope?


#1

Hi,
I’m trying to animate a hydraulic cylinder (red) for a front loader arm (blue). The upper link (green) pushes the cylinder/piston when lifting the main arm (blue).

Q: But how can I keep this push-function (the piston “locked” in the cylinder) while at the same time being able to separately tilt the lower link back and fourth? (I can do each individually, but how combine the two in one mechanism? )

The current model has only a “push-bar” but no “telescope piston”.

Model: test_main_arm 000.3dm (118.6 KB)

// Rolf


(Luc Adriaenssen) #2

Something like this?
test_main_arm 001.3dm (116.6 KB)


#3

Yes, that’s the tilting movement I want.

But I need a simpler way to push/pull the Tilt link since I need to perform very complex lift and tilt movements in my animation.

I tried using Rubberband (but I found no way to control it, I tried change values for scaling, but no “response”). I also tried using telescope,. but same there, nothing happens. I have no idea how to make them expand or retract. And no descriptions in the help file about how they work. Are they simply “followers” driven by other objects? Clueless.

// Rolf


(Luc Adriaenssen) #4

Indeed, an object with joint properties (Hinge, Telescope, Rubberband) cannot be animated directly (via keyframes) in the same motion. So a Hinge cannot be rotated (i.e. installed keyframes will score no result) but it can be moved, whereas a Telescope can be rotated but not moved.
What way simpler than puch/pull were you thinking of. In other words, which element of the structure will cause the piston to function. In my model the upper_piston is simply pulled away from the cylinder by means of keyframes - so the piston itself is the ‘driver’.
Somewhat clumsy of course because the upper_piston’s home position isn’t horizontal, but that can be cured by tilting the entire model.


#5

OK, good to know. I was searching blindly for something ready to go, but no problem, now that I know that no direct functionality is in there, I will make some kind of mechanism that can easily be changed with rotation angle in key frames as to make it possible to tilt quickly and easily along the time line. (video clip shows how far I have come, tilting yet to be fixed).

One more question: I need to go forward and backward along many different paths (I’m to illustrate a driving pattern with some special equipment attached instead of the bucket, but how to change/swap paths? New animation sets for each new path?)

Edit: Trying to render with MovieMaker instead:

// Rolf


(Chris Kuether ) #6

The video is ‘blank’ over here…


#7

Hm, I don’t know why the video fails. Anyway, try downloading it (right click + “Save video as”) and run it with VLC.

Edit: I tried render with MovieMaker instead of Vegas Movie Studio. Hopefully it can play directly in the browser now?

// Rolf


(Chris Kuether ) #8

Yeah. That works. Wonder why/how/etc…


#9

I made a mechanism (blue) to push & pull the piston by applying an angle (+ or - ) in keyframes, this works fine. But I have problems with the links having a too “loose” connection to the bucket. The upper (black) tilt-link against the bucket tends to lose its grip every now and then.

If I only knew of a better (rock solid) “connection” to the bucket, then I could apply this principle on the 3D tractor model.

test_main_arm 004.3dm (212.4 KB)

// Rolf


(Luc Adriaenssen) #10

Very nice idea, the tilt_rotator!!

I made a few alteration in the IK-structures, making it more linear so that less recursions occur.
Here it is: test_main_arm 005.3dm (194.6 KB).
I’m too tired now to write a more detailed explication.

You should take care that parallelograms in this kind of structures are truly parallelograms (equal lengths, equal angles)
I adapted the bucket scissor and the upper_paralell.

Anyway, the bottom line is : it works :sweat_smile:


#11

@Luc,

OK, I made a “final test” based on your cleaner solution, plus an extra link between the Piston and the Bucket (needed in order to flip the bucket very much). But I keep having bad luck (or Luc?) with the “double guided” cylinder/piston sliders.

Watch carefully in this clip, made in three (3) passes, how the main cylinder (lower right corner) typically flips around, in pass 2 and 3 (a very quick flip in pass 3) when the piston end passes the middle hinge.

It seems that when a Constrain To Object-point slides by very close to another point (with Bongo definitions), then the Constrained To Object-point is “disturbed” somehow.

I’ve seen this flipping behavior many times. I tried to “armor” the piston’s “cylinder slider” with an extra To Object point, but to no avail, which indicates a bug. Hopefully this problem can be fixed with a service release.

L70 Bongo Test 012.3dm (520.8 KB)

// Rolf


(Luc Adriaenssen) #12

Rolf,
Try to keep it simple. A simple telescope does wonders.
main_cylinder_cc, pt_main_cylinder_slider, Object 1577, pt_GROUND_C, main_piston_cc, pt_main_piston_slider, can be omitted.
L70 Bongo Test 012 bis.3dm (505.8 KB)

I made 2 minor corrections:
pt_GROUND_B can simple be constraint to its position leaving relax rotation – so there is no need for it to be hinge.
pt_GROUND_A is redundant.

Luc

PS. I enjoy your IK-puzzle :slight_smile:
BTW. I found an even more elegant sliding system for the upper_piston.
test_main_arm 007.3dm (161.5 KB)


#13

Ah, yes, that was a nice one, I’ll keep that. :slight_smile:

Thank you Luc very much for these hints, its been incredibly useful. I now also understand the mechanisms better, which is a prerequisite to know how to do it simple (it is often due to my testing of many variants before getting to work that my test models ends up being a bit über complex… ). But I think I have improved now.

BTW, the reason why I separate the bongo curves and points (a kind of a “exoskeleton” or “stick figure”) from the contours and surfaces is that I can copy that bongo “stick figure” over to my final 3D model and then just re-enter the Bongo defs to make the 3D model live (the contours where copied from the 3D model in the first place, so the structure of joints is exact).

// Rolf