DOUBLE U joint Rigging

[twetzel2]

I have seen and followed multiple single U joint rigs, but cannot seem to get the second half of a double cardan joint to work.  I get a lot of floating or incorrect rotation on the remaining side.  

has any one done this and be able to give me some help?

[leeminardi]

I think the approach I took here could be used.  I would modify the code to handle a normal u-joint (i.e., one without an offset) and where the driven axis points to the second u-joint.

Can you post a Max file of the components?

[twetzel2]

I have attached a file here.   it is a simple blocked out version of what I have. 

[leeminardi]

@twetzel2   please post a Max 2025 compatible version of the file.

[twetzel2]

I have also included my attempt with your hexapod tutorial.  My Joint2 (your Pt B) keeps jumping way off target.

[leeminardi]

Please "save as" 3ds Max 2024.

[twetzel2]

2024 version

[leeminardi]

The Basic double U-joint you have does not include a shaft between the two yokes.   As a result, the relative motion that you will be able to achieve would be similar to the motion in the first half of this video but would NOT support the motion in the second half.  Does that limitation meet your requirements?

I will look more closely at your model later today.

Lee 

[twetzel2]

the video example is correct, but I need to be able to move either end up and down as well.  I don't need the yellow gears.  The motion is not allow, but I do need to have some flexion on both sides.  

The idea is that I would be able to move either end and the joint would adjust.  I am 100% willing to have controls that would help me do this - for example a control that assists the rotation of the center joint.  

[leeminardi]

The rigging I am assuming is that you would rotate the orange component about its centerline axis and the red shaft would rotate accordingly.

You would be able to change the direction of the centerline of the red shaft  as long as its centerline lies in the same plane as the green component and the orange shaft.  If you wanted more freedom in orienting the red shaft you would need something like this:

Note how the connection between the orange and gray yokes allows for changes in distance between the two parts.

Note, that the ...hexapod 2024 file is not compatible with my 2025 Max  Are you sure you saved it in 2024 format? 

[twetzel2]

its actually the opposite.  My red rotates about the y axis.  The opposite end can move around and rotates with the red one. My orange piece, could end up being at angle. See image - the "orange one" is the one in person's hand. 

This image shows the maximum rotation of the joint.  As you can see it is only about 30 or so degrees. 

My red one connects to another single U joint at the other end and its motion would be fairly minimal.  I am not worried about that one.  

Tried again with the 2024 file. 

[leeminardi]

I think that the axis of the component in your right hand has to be in the same plane as the shaft at the top of the photo. 

I will look at the file later today or tomorrow as I got to attend to other issues.

[twetzel2]

here is another attempt.  I have everything working except the rotation of the 2nd joint blue part.  

[leeminardi]

@twetzel2  I looked at your post 13 file and it was not clear to me how to use it.  I tried rotating the tube and nothing happened.  

Here's my latest attempt.  I'm having trouble getting the flange to follow joint 2 since joint2's orientation is a function of the location of flange.  I linked a dummy to joint 2 hoping that I could use a LookAt constraint to orient the flange but this is not possible as it is a circular reference.

Here's one configuration  (I linked a red sphere to the flange to help me visualize the rotation of the flange).

Rotating the red tube yields the following.  Note that the flange is not oriented correctly to joint2.

With the flange selected I gave the ALIGN command and specified that the x axis of flange should be aligned to the x axis of joint 2. This resulted in the proper orientation for the flange.

The challenge now is how to force that last alignment without having a circular reference.  I'll give it some thought.

Here are the transform scripts for the components.

[twetzel2]

Thanks.  When I load yours and redo it for myself, when I move the Flange Yoke  a bunch of skewed scaling happens on the joint 2 and intermediate yoke

[leeminardi]

Please see the attached version.  I've modified the scripts to hopefully avoid creating non-orthogonal matrices.

I think the real-life mechanism requires that the pivot point of the flange lies on the XY plane of tube.

Rotate the red tube in the attached file and give the ALIGN command, x axis and the flange will reorient to the correct orientation. 

[twetzel2]

Ok, thanks for your help.

I appreciate the time you put in.     

[twetzel2 · Accepted solution]

Here is what I have so far. It ended up being a bone based system with IK. 

It works for almost everything except joint 2 doesn't follow the Flange.  I could not get the lookat to work so it kept the rotation correct.  

I tried some basic wiring as well as simple expressions but did not work.

So I have a simple corrective FK control to counter animate that angle.

Please share if anyone has a better solution. 

[leeminardi]

@twetzel2  How do you change the angle between tube and the direction of flange in your rig.

I have something that works fairly well EXCEPT the matrix for joint 2 becomes non-orthogonal at certain angles distorting its shape. To get around some of the circular references I added a dummy that is used for rigging joint 1, intermediate yoke and joint 2.   The direction of the flange is controlled by the direction of line FlangeDirection.  You can edit the location of end vertices to change the angle between tube and the normal to flange.  Just rotate tube about it local x axis.

I've keyframed the rotation of tube in the attached file.

[twetzel2]

Unfortunately the scale is issue is a big issue and so that can't be used.

I control my rig with the 3 orange Point helpers.  The main driver between those is the bone system behind it all and using the IK and lookAt constraints.