Crenellation pattern trough a set of points

Replies to This Discussion

Permalink Reply by Sebastien de Wit on January 20, 2017 at 2:24am

These hinges look funny, could have been an option. Still there are many considerations done, together with my professor. Furthermore the graduation date comes closer and closer, so large changes ain't possible any more. I will discuss such hinges with "false ceilings" at my design studio, perhaps they can something with it :D. 

Looking forward to the more freaky teeth control parameters !!

Permalink Reply by peter fotiadis on January 20, 2017 at 5:19am

Restart (again) from node 0:

Spend a couple of minutes more on that one: get V1D that provides 2 teeth division policies (new variable percentage controls the startOffset in relation with the minEdgeLength [critical for doing the teeth job safely everywhere]). New boolen report variable tells you stories about edges, available length for divisions and the N of divisions.

A small demo thingy gives you a hint about what a double rounded value is: for instance Math.Round(2.49, 0) = 2 whilst Math.Round(2.51,0) = 3). So ... we divide each remaining edge length (length - 2* startOffset) by the new divLength variable (policy == 2) ... we round the result and we get the divisions (min allowed value: 1).  This presupposes that startOffset is constant (should be ??).

And ... well ... the Dark Side wants YOU > don't hesitate to double click on these magic boxes > the logic is exposed (I mean after a proper triple espresso [or more]).

Change anything (even a ".") and have a BIG time, he he.

Attachments:

• CrenelateBrepEdges_V1D.gh, 148 KB

Permalink Reply by Sebastien de Wit on January 20, 2017 at 5:41am

I will study it! Opened the file and this shizzle seems to be perfect :D. I will let you know how it went within a few days (weekend # miller time).

Permalink Reply by peter fotiadis on January 21, 2017 at 1:51am

Restart AND Reset:

1. Forget teeth (only suitable for your mother/father in law), rods (ditto) and the likes.

2. Find a suitable variable aluminum connector (for instance: http://www.fipa.com/en_GB/products/207700-profile-connectors/25/1 ). Find one that can being fixed.

3. Design a custom aluminum beam (or contact Fipa) - BTW. Chinese do custom stuff for peanuts money.

4. Create the vault LBS first using the beams (the "skeleton").

5. Study Migua elastic inserts (critical) and Ceresit PE/S sealants. Get the gist of bridging gaps as a pro.

6. Use marine grade plywood only as a facet top cover (and some proper false ceiling). Plywood dimensions are usually 1.20 * 2.20 m. A 25 mm sheet could be OK for a small vault. DO NOT varnish the plywood. Epoxy glue linear aluminum L (10/10 mm) along the upper lips (in order to allow silicone to adhere properly (not shown in the image below) : failing to do that ... buy an umbrella).

7. Use trigonometry to calculate the variable beam placement per module.

Do this:

Permalink Reply by Sebastien de Wit on January 27, 2017 at 4:40am

Hi Peter, Yesterday I had a meeting with my professor and we were both enthusiast about the results (I will refer to you in my model as well). The other solutions you proposed are an interesting alternative for other graduation studios, but not for mine (I have to stay within the scope of the thesis). So I will start figuring out how to "thicken" the panels so it can be manufactured in February. If you have any suggestions you are always welcome. Cheers !

Permalink Reply by peter fotiadis on January 28, 2017 at 1:09am

Excellent: you both just took the wrong decision (knowledge is only possible via horrible mistakes).

With regard the teeth to teeth puzzle my 2 top parametric gurus suggest:

1. The girl: Use trigonometry

2. The handsome: Use trigonometry.

Permalink Reply by Sebastien de Wit on January 28, 2017 at 6:09am

I suggest option two, but trigonometry is in this case easier said than done. I was thinking to make an offset of each poly line in the direction of each plane. But that can not be done because each polyline has two planes and thus two direction. But if an offset is possible a loft can be created, the resulting loft can then be used as cutter for the plates itself. The only thing that remains after this, is adding teeth at the opposite plate at the location where a teeth is cut away. Sounds logic XD.

Permalink Reply by peter fotiadis on January 28, 2017 at 7:19am

Well ... the handsome guru said:

1. Working on a per Face basis > project each "2 plane" poly on the corresponding BrepFace plane. Assume that we extrude the joined combo both sides at once (meaning that there's clash issues related with the convex edges: the "inward" ones so to speak).

2. If we call the teeth that is contained inside the Face "inside" (Length as in D) and the other "outside" (shorter Length > make a sketch > trigonometry > etc) ... then there's no clash issues ("along" each teeth: i.e. "along" the initial donor edge direction) since the inside meats always the corresponding outside.

3. However there's clash issues related with the start/end portions of the  polyline (due to offset).

4. There's also clash issues ("across" each teeth, i.e. perpendicular to the initial donor edge) .. meaning that the polyline must take into account all these constrains (at creation time ... or at "offset" time)).

Guru refused to dig into more details (God knows what he actually means with all the above). He only added that the trick is an ability to watch BOTH polylines (per adjacent Face) at once (rather easy for him).

Moral: a tipple espresso for me please.  

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page