Autodesk Inventor Design possible in OnShape?

haresh_moradia

Is it possible to create the same type of curve and pattern using OnShape from the video shown here https://www.youtube.com/watch?v=YjoVGhU5hqA&t=95s?

S1mon

It’s certainly possible to model the parts.

What do you mean by pattern? Do you mean the surface texture? That can be done in Renderstudio.

haresh_moradia

What is RenderStudio?  When i create my pattern it results in in tube and out tube going out on angles.  I want both those tubes to be parallel to the XY plane while the loop does it thing and then shoots out back in the XY plane some distance away.  Is it possible to share what I have done so far with you?

nick_papageorge073

Are you after the rollercoaster track shape? If so, you could make a helix curve, then extend the two ends in straight lines. Then sweep the entire curve with the cross section of the track.

S1mon

Post a link to a public document.

haresh_moradia

nick_papageorge073 said:

Are you after the rollercoaster track shape? If so, you could make a helix curve, then extend the two ends in straight lines. Then sweep the entire curve with the cross section of the track.

When i do this the shape intersects with the body but the goal of having the entry and exit are now parallel.  Also, when they go into the curve, you can see how the track is jagged into it, not a smooth transition.

martin_kopplow

If the run-in and the run-out are to be straight and parallel and perpendicular to the helix axis, you'd either need a helix tool that can flatten the ends (like in a cylindrical spring) or you'd have do do that yourself, but in any case, if the tracks in the helix are to touch each other, there will be a self intersection at the ends. No way to avoid. You could leave a space in between, of course. I guess the tool used for your sample doesn't care for self- intersection. Or you could make the run-in and run-out not straight but curved only in their middle part. That would eliminate self- intersection, if done right. Or you could make the loop in two parts that are non-self-intersecting each, and then boolean them, accepting the self-intersection. 

Post a link to a public document to examine.

Here's mine, with a smooth transition between loop and straight rails: https://cad.onshape.com/documents/371ec8e328f0a1e187d39f00/w/d5df61e2525fe93ee6a5cb16/e/c07b615f09d309eabadff19a?renderMode=0&uiState=65568a01f0eca15001f88db0

Edit: After watching the Vid another time: There is a gap and the straight tracks are neiter parallel nor rectangular to the helix axis. There is no secret spice in that model, except for the part where he explains how beginners select the wrong sketch plane for the track profile.

haresh_moradia

martin_kopplow said:

If the run-in and the run-out are to be straight and parallel and perpendicular to the helix axis, you'd either need a helix tool that can flatten the ends (like in a cylindrical spring) or you'd have do do that yourself, but in any case, if the tracks in the helix are to touch each other, there will be a self intersection at the ends. No way to avoid. You could leave a space in between, of course. I guess the tool used for your sample doesn't care for self- intersection. Or you could make the run-in and run-out not straight but curved only in their middle part. That would eliminate self- intersection, if done right. Or you could make the loop in two parts that are non-self-intersecting each, and then boolean them, accepting the self-intersection. 

Post a link to a public document to examine.

Here's mine, with a smooth transition between loop and straight rails: https://cad.onshape.com/documents/371ec8e328f0a1e187d39f00/w/d5df61e2525fe93ee6a5cb16/e/c07b615f09d309eabadff19a?renderMode=0&uiState=65568a01f0eca15001f88db0

Edit: After watching the Vid another time: There is a gap and the straight tracks are neiter parallel nor rectangular to the helix axis. There is no secret spice in that model, except for the part where he explains how beginners select the wrong sketch plane for the track profile.

Interesting!  If the angles are at 90 degrees it seems to do the trick.  I chose an entry angle of -7.5 from vertical and then a launch angle of 30 degrees.  If you implemented those angles in your design, what does it look like.  Thank you so much for taking the time to demonstrate in OnShape.  I think it may be related to my entry and exit angles.  I am curious what your design would look like if you modify those.

martin_kopplow

@haresh_moradia: It does in fact not depend on entry or exit angles. Any old angle will do. You just need to use a different set of sketch planes, then, to start the transition and also to place the profiles. In my model, I used the standard planes, because I had no intent of modifying the angles later, else I'd have used sketch planes, which move with the helix angles. Then, one could modify the whole model by just typing different numbers into the helix properties dialog. (See updated document for an example of helix-driven sketch planes.)

Give it a try and let us see what you get out of it. :0)

haresh_moradia

martin_kopplow said:

@haresh_moradia: It does in fact not depend on entry or exit angles. Any old angle will do. You just need to use a different set of sketch planes, then, to start the transition and also to place the profiles. In my model, I used the standard planes, because I had no intent of modifying the angles later, else I'd have used sketch planes, which move with the helix angles. Then, one could modify the whole model by just typing different numbers into the helix properties dialog. (See updated document for an example of helix-driven sketch planes.)

Give it a try and let us see what you get out of it. :0)

Thanks for this update.  I wanted to share a picture here but it did not allow it.  But if you look at the the two designs you have and look at from the RIGHT.  notice the one on the left is not parallel, but the one on the right is.  But notice how the enter and exits appear to go off on angles?  The reason i am trying to make it planar is so that i can mount the design to a board after printing it out.  I am sharing my actual design which is what we are working on.

martin_kopplow

@haresh_moradia Hi Haresh, yes, my design explores different ways to deal with the angles of the entrance and exit 'ramp' or whatever we may call them. The older one (The left one, when in RIGHT view) approaches that by conically reducing the helical pitch of the rail to zero in the first and final 90°. That makes the exit/entrance rails somewhat parallel to the world planes, at the cost of a possible self intersection. The helix also twists the profile a bit, which is compensated by twisting the entrance rail slightly in the opposite direction, so the starting point is parallel to the world, again. The exit ramp is parallel anyway.

The newer example was just me playing around with the definition of start and end angle. Both ends are not in parallel with the world. Would we extrude them in a tangent of the helix at the end point, they'd be slightly biased, and should we extrude them in parallel to Plane 6, they would create a bump in the marble run groove.

Now I have seen your design (and hopefully understood the intent) and played with my examples a bit, I'd probably start a third approach in a different way,  simplifying the setup quite some:

I'd start out with a tube, easily mounted to the board you mention, and the inner diameter driving the diameter of the helix. I'd then only drive the groove by the helix and create a boolean for the the run-in and run-out, driven by the start and exit angles (or the helix end points and tangents). That minimizes the amount of surfaces which need to be taken care of and it would streamline manufacturing quite some. 

See how many small non-tangent faces and you have in your design? These are not needed for the task at hand and would just not happen if the design started with a clean and simple tube.

I wonder if the whole thing could even better be designed from run-in, run-out and the helix portion in between, all being designed in context on sketches placed in an assembly, with variables driving helix diameter, helix pitch, entry angle and exit angle.

haresh_moradia

Thank you very much for this information.  I am having some hard time to understand how you mean to design using the easy approach you describe above.  You are correct, that this is for doing a marble run.  I see your point about all those triangles.  How to make a nice smooth curve that would allow a marble to make a loop to loop and transition without turning along the curve.  That has been our challenge.

haresh_moradia

Do you have another example that you can share?

martin_kopplow

@haresh_moradia Well, I've got one more for you. As you can see, it has a much cleaner shape, with almost all faces being orthogonal to the mounting surface, except for the loop groove and the lower portion of the run-in track. 

In this sample, I have only modeled the run-in, but the run-out could of course be made in the same manner. Maybe you wanna givvit a try yourself?

This model is based on a simple cylinder, and it has two helixes: One takes care of where the loop groove is, and the other one takes care of where there is no more loop, because it has to end somewhere, depending on the run-in angle. There is one important plane that drives it all: If you edit the angle of "Plane Run-in Angle" within reasonable range (start with say 20° or 30°, I haven't tried how far you could go), the run-in angle will adjust while at the same time the cylinder that serves as the loop will adapt its length and the groove start point to the new angle, while the end point remains fixed (until you model something that adjusts it). 

https://cad.onshape.com/documents/87a605abcfacbedaa9f35736/w/91153fdbec95cc5994b8e62e/e/d76d158e1b0f793ac65d4c02?renderMode=0&uiState=655fdf30184f9525d2dc5647

I have accidentally left one small bump where the run-in attaches to the lop, though. That is because I was a bit inaccurate with one sketch, ant there is certainly room for improvement, but since it is already half past midnight, I'm going to leave it like that for now. Try to model the the run-out and have fun!

haresh_moradia

martin_kopplow said:

@haresh_moradia Well, I've got one more for you. As you can see, it has a much cleaner shape, with almost all faces being orthogonal to the mounting surface, except for the loop groove and the lower portion of the run-in track. 

In this sample, I have only modeled the run-in, but the run-out could of course be made in the same manner. Maybe you wanna givvit a try yourself?

This model is based on a simple cylinder, and it has two helixes: One takes care of where the loop groove is, and the other one takes care of where there is no more loop, because it has to end somewhere, depending on the run-in angle. There is one important plane that drives it all: If you edit the angle of "Plane Run-in Angle" within reasonable range (start with say 20° or 30°, I haven't tried how far you could go), the run-in angle will adjust while at the same time the cylinder that serves as the loop will adapt its length and the groove start point to the new angle, while the end point remains fixed (until you model something that adjusts it). 

https://cad.onshape.com/documents/87a605abcfacbedaa9f35736/w/91153fdbec95cc5994b8e62e/e/d76d158e1b0f793ac65d4c02?renderMode=0&uiState=655fdf30184f9525d2dc5647

I have accidentally left one small bump where the run-in attaches to the lop, though. That is because I was a bit inaccurate with one sketch, ant there is certainly room for improvement, but since it is already half past midnight, I'm going to leave it like that for now. Try to model the the run-out and have fun!

AMAZING!  This is awesome.  Now to ask another question, i want the 10 degree angle to be placed on the back side so that on the front you can see the 30 degree launch angle.  Is this possible?  Thank you for these wonderful ideas and tips.

martin_kopplow

Hi @haresh_moradia , yes, that is of course possible. You could build the model both directions, or both at a time, if you want. Just start with an anti-clockwise helix, where I used the clockwise direction. That will put the run-in track near the wall mount, and leave you with the launch track at the front. I am not an AI bot, though, so I am afraid you will have to do it yourself to also earn the benefit of learning it yourself. Who knows? You might even find a much better solution.

Well, you could of course just mirror the thing and enter -30° as the angle, but that still wouldn't do the magic trick to make the now missing run-in suddenly appear from thin air. Feel free to post your result, if you have further questions.

haresh_moradia

I will give that a try right now to see what happens

haresh_moradia

Well this didnt go too well

haresh_moradia

Thanks for the motivation and advice...check this out...any feedback would be awesome!



Not very pretty but working at it.

martin_kopplow

Hi @haresh_moradia

nice approach with the piped-in track! You could use that at both sides and even include the angular transition in the first tangent part. Referenced to a point on the helix, it should give you a fully configurable loop with run-in and run-out.

In the previous post, did you use -30° or 30°? For me, -30° does work, though the model will collapse at more than that. It wasn't initially built for this, though.

Anyway, you're on the right track, keep up the good work.

haresh_moradia

Thank you again for the encouragement.  The plan is now to automate this using variables to easily make changes depending on steel marble size

martin_kopplow

Great! I'd really like to see the result. I'm admittedly a big fan of fancy marble runs.

haresh_moradia

absolutely!

haresh_moradia

martin_kopplow said:

Great! I'd really like to see the result. I'm admittedly a big fan of fancy marble runs.

It was a success!!!  We were able to change the diameter and print them out.  Now we are just optimizing the distances launched.

martin_kopplow

@haresh_moradia Congratulations!

haresh_moradia

martin_kopplow said:

@haresh_moradia Congratulations!

Thank you again for sharing your insight and approach on how to solve the problem.  The kids understood what you described much better than myself and in the end they were able to print their models.

martin_kopplow

Super! Let us see a picture when it's ready. Everybody likes marble runs, right? :0)

haresh_moradia

martin_kopplow said:

Super! Let us see a picture when it's ready. Everybody likes marble runs, right? :0)

We should have a picture for you tomorrow!!!  And then shortly a video.  We do have another question.  They have a design that they want to make like a funnel but to make the exterior solid like a cone so that they can mount it.  How can I share the design that they created to help give us how to make this solid and mountable?

We want to make the above helical a solid but empty in the middle so that we can mount this.  

martin_kopplow

That should be possible as well, and certainly fun to watch the marble circling ever faster as it goes down the funnel.

To share a design, follow step 2 in this post:

https://forum.onshape.com/discussion/9107/forum-post-checklist-read-this-first#latest

haresh_moradia

https://cad.onshape.com/documents/3f38e115df7e9c6c968e4133/w/4e25c35a134864e537eaf6ff/e/8eece2f69cc288ce3e896c22

martin_kopplow

Great, that works! We can now see the file.

The issue with your solid funnel appears to be the spiraling track cannot self-intersect.

In this case, I'd try designing the funnel wall first, then create a sketch of a 'negative' track cross section, sweep that along a spiral that lies on the funnel's inner surface and boolean-subtract that from the funnel. You already did that, though you modeled the inner body of the funnel, not the outer wall. I'd just swap the whole thing.

Here's a sample: I left everything in place and only added a folder with what I did:

https://cad.onshape.com/documents/64eae8c4209a02df363de795/w/24d7768046cc8cedab69efed/e/ca97d8af2dae0cc4f8f3b85d?renderMode=0&uiState=65bea5165a26f6028e2dca3e