Circular Transition

martin_kopplow · March 12

I have a seemingly simple problem. I want to create a smooth transition between the 2 slabs. It is to be tangential at the start and end profiles, and that is easy so far. But I also want it to be circular in the plan form. Now, I could play with the start and end magnitude of the loft, until I get pretty close, but this provides little control, modifies the slope at the same time, and then I'd like a precise method, one I could reproduce exactly each time, and possibly also make configurable later.

Any ideas (except for replace face, which does neither work nor is it precise) for this simple one?

Image: https://us.v-cdn.net/5022071/uploads/editor/ww/b10om699d6y9.png

https://cad.onshape.com/documents/6f6fe0b0a851c362ced79d4f/w/cc419924a48e7ddf35a00be1/e/804624dc8af58e762704fc52?renderMode=0&uiState=65f0b68764ab0a17afbf6085

MichaelPascoe · March 12

One way would be to draw guides on the left and right side via sketches, then project them to the cylindrical surface. Then loft the two guides together, or use them as guides for the one solid loft.

https://cad.onshape.com/documents/45f3e544fd918a950b6489e6/w/cb2f77229993baa59e9daf2b/e/f276d2...

martin_kopplow · March 14

That would be a possibility, even though the projected slope gets a bit of distortion towards the ends. The advantage here is that it creates a boole-able result at the end, which the wrapped-around sketch method didn't.

jelte_steur814 · March 14

Hi Martin,

the surfacing option would be to loft or boundary surface the top and bottom faces, and then mutual trim the sides.
like so:

https://cad.onshape.com/documents/8f0dc9c64a802eb2b3e126ef/w/c2fdf4bed38e57c3c96ed60e/e/e9300ab1942186f21e656232?renderMode=0&uiState=65f327cdc1b8263c21e0bfe0

the tangency_magnitude parameter can be used to manipulate the smoothness for all 4 transitions in one go.
so in terms of smoothness i think it probably may match your expectations.

I'm not sure this would work robustly in a configurable setup though. the 'move boundary' is rather ugly and also mutual trim directions may flip and cause errors.

kind regards
Jelte

martin_kopplow · March 14

Hi @jelte_steur814 yes, looks good. That's a few more steps, but it creates precise geometry in the right places. Maybe I could use split by curves from the bottom sketch as an alternative to trim. I'll give that a try.

STEG · March 15

Hi @martin_kopplow,

You could create 4 Bridging curves (with Match tangent option) between the corners and then use 4 Boundary surfaces (with Boundary condition set to Match tangent for V profiles). Note that all dimensions are random for my 2 slabs.

Like this:
https://cad.onshape.com/documents/5cb32d69b83de94b79486e00/w/bd9cfbd6a34092b4703880fd/e/f777f6752915e445a0f46523?renderMode=0&uiState=65f443a808dd463894cdc15f

Image: https://us.v-cdn.net/5022071/uploads/editor/l1/touny9bwg723.png

martin_kopplow · March 15

Hi @STEG

sure, but the briding curves won't create a quarter circle, they just make a more or less random shape in plan view, don't hey? So this creates the same shape a simple loft from the solid slabs would, it just takes more steps to set up. I need to precisely control the vertical faces.

STEG · March 15

@martin_kopplow sorry didn't note about the quarter circle... How about this workflow?

https://cad.onshape.com/documents/5cb32d69b83de94b79486e00/w/bd9cfbd6a34092b4703880fd/e/fc281499b0e98d03c567d1a2

Image: https://us.v-cdn.net/5022071/uploads/editor/1x/crrjbp86rld0.png

martin_kopplow · March 15

@STEG yes, that's a bit like Michael's approach, only he used a sketch plane at 45°. So we have more than one way to skin a cat, obviously. :0) That's great:

Image: https://us.v-cdn.net/5022071/uploads/editor/23/24i4ev2ubpqd.png

Wrapping a sketch provides most control over the slope, but has precision limits where connecting to the other slab.

Image: https://us.v-cdn.net/5022071/uploads/editor/tw/gfgwbvmsrucj.png

Projecting at 90° creates a proper connection to both slabs, but can't be controlled regarding plan form.

Image: https://us.v-cdn.net/5022071/uploads/editor/oy/ugpi3zorc8ns.png

Projecting a sketch at 45° can be controlled both directions and makes good connction to both slabs. A radial section through the axis clearly makes a curve, though. That means the transition surface can't be made from sheet material.

Image: https://us.v-cdn.net/5022071/uploads/editor/b0/wdp5owluaih2.png

So far I got to this point, where a screw-shaped loft is made with some excess width, then the circle segments from the plan sketch are used to trim the lofted face. Connection to the slabs is good, planform is a precise circle, and the radials are very close to a straight line (though not 100%, blame the clash between tangent end condition and vertical guide curve), so that I could have a flattened shape cut from a sheet within reasonable tolerances and get it mounted on a substructure designed based on this shape.

https://cad.onshape.com/documents/6f6fe0b0a851c362ced79d4f/w/cc419924a48e7ddf35a00be1/e/804624dc8af58e762704fc52?renderMode=0&uiState=65f47d8d037bc110b51e31d0

At this point, I am running out of ideas. I was also thinking about using a quarter helix, but then, Onshape would need a helix tool with closed (parallel) ends, to make a clean tangent connection to the slab surfaces. This is a situation, where Catia can obviously do more, for there, I could fully define a sloping curve onto a cylinder face. Could feature script do that?

ilya_baran · March 15

Was curious enough to try this -- does this work? https://cad.onshape.com/documents/48e11603f915926784c90780/w/a998c8d191153d47849d2091/e/70f244c8fb1e05514f49c1e3

S1mon · March 15

I had to try too. Looks like I did something similar to @ilya_baran at the same time.

Image: https://us.v-cdn.net/5022071/uploads/editor/wm/zc09vt88lcw4.png

https://cad.onshape.com/documents/3af4454171b98709f4c7eb7a/w/717b3fdedf1ada614af0091f/e/cf1aa36c605692c00bf47bce

MichaelPascoe · March 15

Ah, the wrap is a great idea. This will give you exact control. Since it is a simple 90 and a cylinder, this should work fine.

martin_kopplow · March 16

Yeah, the lock face sweep ia also a nice idea. It is pretty new, so I did not have it on my radar: A good combination of control over the end conditions and slope control.

@MichaelPascoe: The wrap has the disadvantage of not creating a precise end. It would offer maximum configurability (or how would one call that?) because the sketch could be fully built on variables or re-used vertices from the slabs, and the radis and angle drawn from the plan sketch. I connected the lenght of the wrapped sketch to the arc lenght, to make it precise, but it didn't come out as precisely as needed to get a non-intersecting match with the second slab. Probably a question of how many decimal digits are available. If I could overcome that, it might become the method of choice for the project at hand.

steve_shubin · March 16

martin_kopplow said:

Any ideas ?

Maybe. It depends.

This is NOT some type of FOOT PATH. Right ?

MichaelPascoe · March 18

Lol, this is going to be one legit sidewalk transition! I want each of our user names etched into this when you finish.

@martin_kopplow, the wrap approach can be achieved by measuring the length of the inner and outer radius, then constraining the heights of the wrapped sketch splines horizontally to the vertices of their end locations.

Here is an example of precision wrapping:
https://cad.onshape.com/documents/45f3e544fd918a950b6489e6/w/cb2f77229993baa59e9daf2b/e/892acec2dec20d87e0ff2...

martin_kopplow · March 18

No, guys, this is NOT a hyperprecise sidewalk. :0)

Should I get the contract, I'll show pics after it has been installed. You'll know what it is, once you see it, and it's going to be on the internet anyway ;0)

@m@MichaelPascoe I only wrapped the outer cylinder, using the same method. Could be wrapping the other cylinder as well enhances precision. I'll give it a try. Thanks for the input everybody!

steve_shubin · March 18

martin_kopplow said:

No, guys, this is NOT a hyperprecise sidewalk. :0)

Great to hear that !

For what it’s worth - for any interested in pedestrian ramp design, here is a good starting point.

https://www.ada.gov/law-and-regs/design-standards/

But I do like all the G1, G2?, and G3? continuity. Wonderful stuff.

martin_kopplow · March 18

And I feel very safe now, knowing that the 1991 ADA Standards were adopted in 1991, whereas the 2010 ADA Standards were adopted in 2010. Good they wrote that down once and for all times. I thought german regulations were weird, but in comparison, this is pure abstract poetry! When I grow up, I now want to be a sidewalk artist!

steve_shubin · March 18

martin_kopplow said:

this is pure abstract poetry!

LOL

Yeah, those government people do like to get wordy that is for sure.

Here’s one with a bunch of illustrations

https://www.access-board.gov/files/ada/guides/ramps.pdf

Circular Transition

Comments

Howdy, Stranger!

Quick Links

Categories