Trouble with rotation translation with a roller between two planar surfaces

[dtaylorZ9JTS]

Hello, All.  What I want to do is model a surface which is fixed, a cylinder which rolls along that surface, and another surface which rolls along the surface of the cylinder.  I've worked on this for hours and tried a number of example solutions but I've not found an answer which works yet (at least the way I want to model it).  As I proof-read this before posting I will put this here at the beginning which duplicates it later: think of a board rolling on a pipe on a table.  The table doesn't move but the cylinder and the board do move.  

In the attached example files: I can get the Red Bar to work properly with the Wheel.  I can lock the Wheel's axis to a hole on the Plate (unsuppress "Suppress B1"), suppress the Flush constraint on the end of the Red Bar ("Suppress A1"), and set Rotation/Translation: Red Bar to "Reverse" and Rotation/Translation: Green Bar to "Forward", then dragging the Red Bar causes the Green Bar to move as I expect.  That is, dragging the Red Bar one direction causes the Green Bar to move the opposite direction and the same distance the Red Bar is dragged (which effectively means the Green Bar moves twice the distance the Red Bar is moved from the Red Bar).

What I would like is to be able to lock the Red Bar's position then drag the Green Bar and have the Wheel roll between the Red Bar and the Green Bar.  To attempt to achieve this, I unsuppress "Suppress A1" (flush constraint on the end of the Red Bar), suppress "Suppress B1" (the axial constraint on the Wheel to the hole in the Plate).  This locks the Red Bar in place and allows the Wheel to travel along it.  I set the "Rotation/Translation: Red Bar" to "Forward" and dragging the Wheel causes it to roll along the Red Bar as expected (movement/rotation appears correct).

This is where my problem lies: If I set Rotation/Translation: Green Bar to "Reverse", dragging the Wheel along the Red Bar causes the Green Bar to move what appears to be the correct distance but *not* the correct direction.  When I change the direction of Rotation/Translation: Green Bar to "Forward", dragging the Wheel along the Red Bar causes the Green Bar to move the correct direction *but* the Green Bar retains it's position in relationship to the Wheel.

What I'd like to be able to model is this: a planar surface rolling on a cylinder across another planar surface (like a board rolling on a pipe across a table).  The extents of motion would be when the cylinder reaches either edge of either planar surface (that I can achieve with a Translation constraint which I left off this example to not overly complicate matters here).  I'd like to be able to drag the "unlocked" surface (the "board" in this example) and see the motion and extents of motion. 

I hope I've done a good job of explaining what I'm trying to achieve and the difficulties I've had.  The attached example should help greatly.  I've named components and constraints in the assembly as they're referred to here.

Truth in advertising: I totally stole an example file someone else posted and used the components from it.  I found that example file while searching for a solution before posting this.

Thank you.

[kacper.suchomski]

Hi @dtaylorZ9JTS 

First thing - reverse the rotational constraint between the circle and the green bar. Currently turning the wheel causes the green bar to move in the wrong direction.

[kacper.suchomski · Accepted solution]

The second thing - if you block the red bar; set trade value circle-green bar x2.

This will result in the correct proportion of work.

[kacper.suchomski]

Now the mechanism works when the wheel is moving.

All constraints seem to be fine.

I think it might be a bug in the program - the fact that this binding only works one way; contrary to what the manual says:

https://help.autodesk.com/view/INVNTOR/2023/ENU/?guid=GUID-0293C151-565F-4980-AA95-DD76C54AB538

It can certainly be done in a dynamic simulation environment. But that's another level.

If no one else on the forum finds a bug in the model, you can report a bug in the software; here or at Idea Station.

[CCarreiras]

---

@kacper.suchomski wrote:

The second thing - if you block the red bar; set trade value circle-green bar x2.

This will result in the correct proportion of work.

---

HI @kacper.suchomski !

Why set trade value circle-green bar x2  ??

[kacper.suchomski · Accepted solution]

Hi @CCarreiras 

We have this situation:

1. Translation between the red bar and the wheel

2. Translation between the green bar and the wheel

3. Grounding the red bar (via constraints)

This causes - when you spin the wheel it makes 2 moves. It rotates and travels a distance equal to its circumference.

Then the green bar must follow the normal route (resulting from the gear ratio) + the one resulting from the wheel displacement.

I am confident that in a dynamic simulation environment, Inventor would understand this without any additional settings or overrides.
But as you can see, constraints have their functional limitations.

Have I been able to explain?

[CCarreiras]

Ok ... trans + rot.... x2... make sense.
Well done, Thank you!! 

[kacper.suchomski]

This constraint only considers linear-rotational relationships between the elements it includes; but he does not look at the model as a complete mechanism. Does not include mechanical friction etc.
This makes it difficult to work - you have to anticipate the limitations of the program and the consequences of your choices; but no software is perfect.

But we still do not know the answer to the author's main question.

Why doesn't this constraint work both ways, as the developer assures?

[dtaylorZ9JTS]

Kacper, thank you for your reply.  I will answer your posts in order for clarity.

For this response, yes, the rotational translation is going the wrong way as I mentioned in my post.  The reason for allowing this to remain is to show that the rotational translation will move the Green Bar an amount we'd expect but just the wrong direction.  If we change the rotational translation direction, the Green Bar doesn't move in relation to the Wheel.

Again, I will respond in order so your next point will be addressed in a separate reply.

Thank you again for your time, effort, and help.

[dtaylorZ9JTS]

Kacper, thank you for this reply.  Changing the translation distance to 2x solved the problem (once I changed the translation direction (which I knew was wrong but the "right" direction didn't produce any movement between the Green Bar and the Wheel) which you also pointed out). 

It is strange that the translation doesn't keep relationship between the Green Bar and the Wheel correctly.  One would expect that if the Wheel moves X units in relationship to the Red Bar when turned, and the Green Bar moves X units in relationship to the Wheel when the Wheel is turned, that would effectively move the Green Bar 2X units.

If it's a bug in the software it's understandable.  Jeez...the incredible amount of tasks and functions this software performs is mind boggling.  It is a hinderance, though, as the modeler will need to be aware of this and account for it.  Just one more "task" to add to the load.

Thank you again for your time, effort, and help.  I have a follow up question but I'll post that as a response to my original question for clarity.

[dtaylorZ9JTS]

Ok, Kacper solved the main problem I was having (thank you).  Now, the follow-up question:

I changed the constraints in my model and got the Green Bar and Wheel to move as expected (direction and distance).  I get the model to function by dragging the Wheel.  It would be nice to be able to drag the Green Bar and have it move the Wheel along the Red Bar as the mechanism would in real life.  As it is now, if I drag the Green Bar it moves independently of the Wheel and Red Bar.  In another model using these components, if the Wheel is constrained by axis to the hole in the plate, direction of the translation is changed, and I suppress the Flush constraint from the Green Bar to the Plate, I can move the Green Bar and have the Red Bar move along the Wheel as expected.  This also seems to be inconsistent in the program.

I'll attach that file set as well.

Thank you.

[kacper.suchomski]

Yes. I mentioned this problem in one of the comments.
I checked the stable wheel and it works for me too.
It seems that Inventor can't handle so many degrees of freedom and relations at the same time.

[JDMather]

@dtaylorZ9JTS 

Extract and open the Attached assembly.

Go to Environments>Dynamic Simulation.

Click on Play

You can also drag the wheel or green bar with mouse - but if you drag faster than Inventor can do the calculations you will get an error.

[dtaylorZ9JTS]

Kacper, I'm sorry I missed that comment then.  I agree about the degrees of freedom.  In my many attempts before posting for help, I saw behaviors change depending on constraints being applied to some components.  Not exactly intuitive, either.  I can't reproduce them, though, as I was banging on a solution...no real finesse to my efforts I'm afraid.

Thank you.

[dtaylorZ9JTS]

JD, thank you for your reply.  You're a huge help on this site and I see you now do YouTube videos as the CADWhisperer.  Lots of respect and appreciation for your knowledge and efforts to help others.

I loaded the file and ran the sim successfully.  However, forgive me, I feel daft: outside the simulation, dragging the Wheel or the Green Bar only moves the individual piece.  While the sim is running, attempting to drag a component only pauses the sim temporarily while the mouse button is down.  While in Dynamic Simulation, but with the sim *not* running, dragging either the Green Bar or the Wheel does nothing.  I'm sorry if I'm doing something stupid or just not grasping the concept somehow.

Thank you again for your reply.

[JDMather · Accepted solution]

---

@dtaylorZ9JTS wrote:

While in Dynamic Simulation, but with the sim *not* running, dragging either the Green Bar or the Wheel does nothing.

---

@dtaylorZ9JTS 

There are two methods to animate:

1. Click Play on the Simulation Player, (Click Rewind after stopping and Play again will run faster as the calculations are done).  Click Construction Mode when done.

2. Click and drag with mouse.

The black arrow represents a Force Vector.  The longer and bigger the arrow gets - the more force that is applied.

You have to hold the vector for a second or so to start the motion - it isn't instantaneous.  Once motion start reduce the Force Vector or it will quickly run up in velocity and perhaps overrun the calculations of motion.

Increased Force...

Note that I am not going parallel to travel - so Inventor is calculating the forces not parallel to travel too.

That is why it might be slow.

If I apply the Force Vector more or less parallel to direction of travel it solves faster but can quickly move too fast to keep up the calculations - reduce the force once moving.  Just like the real world.

It is always most reliable to use the Player rather than manually drag.

If desired - you can then Publish the simulation to Studio.

[dtaylorZ9JTS]

JD, thank you for your reply.  I see what I was doing wrongly now.  I had tried the Construction button because that's how I am: click/bang on everything.  I had tried to drag the bar and had seen the line appear but no motion take place.  I now understand that the line I was seeing was a "force" line and, yes, the movement is not instantaneous.  I'm also pretty fast and I didn't see any movement (yesterday) and moved on.  That caused me to miss that what I was doing was affecting something.  Today, after reading this last post, I tried it again.  Yes, as you said, it works.  Dragging the bar works.

Thank you again.  Sorry, I didn't know boo about how that feature/interface worked.