SDL with placeholder values

Replies to This Discussion

Permalink Reply by Joseph Oster on June 10, 2017 at 4:20pm

Well...  First of all (as I think you know), an 'SDL' line is defined by Start, Direction and Length, not by an end point.

Second, is an 'SDL' really necessary at all here?  Maybe just connecting two points will work?

Third, though I've re-read your descriptions, I can't quite figure out what you want to do?

I started to write something, then noticed that the two circles aren't on the same plane (not obvious from the 'Top' view), so changed from a circle intersection to sphere intersection.  Are you thinking of something like this?

Attachments:

• pushrod_sample_2017Jun10a.gh, 11 KB

Permalink Reply by Ethan Gross on June 10, 2017 at 4:24pm

If I understand you correctly, I think you need to solve for the intersection of two circles, one with radius = length of the pushrod and one with radius = length of the "follower" or whatever it's called. SDL won't work here, at least not easily.

Attachments:

• pushrod_linkage.gh, 10 KB

Permalink Reply by Ethan Gross on June 10, 2017 at 4:40pm

Until I saw Joseph's interpretation I didn't realize the two circles were in different planes, so here is my new solution, same as the old solution except now it's based on the intersection of a sphere with a circle. Of course, this still assumes I'm reading this correctly...

Attachments:

• pushrod_linkage2.gh, 13 KB

Permalink Reply by Joseph Oster on June 10, 2017 at 4:53pm

OK, but as noted, his circles aren't on the same plane so a sphere is needed instead of a circle.

'List Item' always returns the first of the two points, which causes a problem when they suddenly reverse.  I "fixed that" by using 'AlongCrv (Sort Along Curve)' before passing the two intersection points to 'List Item':

Attachments:

• pushrod_sample_2017Jun10b.gh, 15 KB

Permalink Reply by Ken Charron on June 10, 2017 at 7:33pm

Gentlemen. Just getting back now and already some very innovative thinking.

About a week ago I did have something in the middle as a start point for a sphere. What I made allowed a sphere intersection (as a brep object) to intersect with an ARC (rather than a circle).

Yes I intentionally made the circles at different elevations. This is because the two circles (ARCs) in my project have different plane angles. So the end solution needed to be independent of a 2D model.

What failed in my earlier attempt was the fact that my sphere at the centre of the lines (because I used two lines, each going towards separate ARCs) the lateral sphere movement I so desired control over meant that the path between each circle could never be straight, so the line between ARC points always had a kink thanks to the sphere not sitting in between (it harder to do than you think).

In real-world terms, what I’m looking to achieve are two wheels with a pipe connection between them — like a steam locomotive. The tricky part is, you can only apply lateral movement to the connecting rod. You cannot apply a rotational force to one circle.

Ethan. The sample you have there is where I was three weeks ago. The "move me" slider is an arbitrary input. Were it an input in degrees, that would really be something, but sadly it doesn't allow for that.

Permalink Reply by Ken Charron on June 10, 2017 at 7:43pm

Jeff Hammond over on McNeil sank this ship. Check out his trig solution...

Attachments:

• pushrod_jh1.gh, 14 KB

Permalink Reply by Ken Charron on June 10, 2017 at 8:05pm

I knew it was too good to be true.

Seems this arrangement only allows the final line to be parallel with the centre of both circles rather than allow the line to maintain a fixed length while sliding along the path of each circle (using a lateral input slider).

Permalink Reply by Ethan Gross on June 10, 2017 at 8:05pm

Why don't you specify what you're looking for next time? If you wanted a trig solution, then you should have asked for it. Now that would have really been something!

Permalink Reply by Joseph Oster on June 10, 2017 at 8:18pm

I think you're making this sound much harder than it really is:

Attachments:

• pushrod_sample_2017Jun10c.gh, 17 KB

Permalink Reply by Joseph Oster on June 10, 2017 at 8:44pm

The only thing tricky about this is setting the position and radii of the two circles such that the "rod" isn't too short or too long through 360 degrees of rotation:

Attachments:

• pushrod_sample_2017Jun10d.gh, 25 KB

Permalink Reply by Ken Charron on June 11, 2017 at 6:53am

Holy Cow!

I think I will need to poke through this for a bit.

Amazing!

Permalink Reply by Joseph Oster on June 11, 2017 at 7:11am

Again, this really isn't that complicated.  Here is a slightly simpler version:

Try by-passing the 'AlongCrv' component and see what happens.

P.S.  Hey, wait a minute!  'AlongCrv' doesn't seem to have any effect in this version!?

Attachments:

• pushrod_sample_2017Jun11a.gh, 21 KB

• ‹ Previous
• 1
• 2
• Next ›
• Page