Tesselion : Adaptive Quadrilateral Flat Panelization

Just another WordPress.com weblog

More Final Images

with 4 comments

Tesselion is located at the Philadelphia University Architecture & Design Building & will be up until October.

Address : 4201 Henry Ave. Philadelphia, PA 19144

Feel free to visit & take photos (send me any images & I will post them on the web)!

Written by sjet

May 16, 2008 at 7:52 pm

Posted in Uncategorized

Tesselion Complete!

leave a comment »

Images from the completed installation,

Tesselion: Adaptive Quadrilateral Flat Panelization.

www.tesselion.wordpress.com

Material Sponsor: Alliance Metals ( www.alliancemetals.com)

Fabrication Sponsor: Jared Laucks & Continental Signs ( www.continentalsigns.net)

Many Thanks: Jared Laucks, Marc Fornes, Adrienne Yancone & everyone else for the generous support!

More images soon!

Written by sjet

May 8, 2008 at 6:49 am

Construction Day 17_5:4:08

leave a comment »

A few photos as the canopy touches down with the ramp. 

Assembly has been completed: (Final photos will be up soon!)

-All panels are in place & tightened with lock nuts (except 1 which needs to be cut & shipped to site).

-The entire construction including folding, rivetting, bolting, installing & tightening took roughly 7 (12 hour) days for with only 1 person.

Written by sjet

May 5, 2008 at 6:04 pm

Posted in Construction

Construction Images Day 15_5:02:08

leave a comment »

Written by sjet

May 4, 2008 at 1:47 am

Posted in Uncategorized

Construction Images Day 14_5:01:08

leave a comment »

Written by sjet

May 4, 2008 at 1:44 am

Posted in Uncategorized

Construction Images Day 10 : 4:27:08

leave a comment »

The threaded rod has been set into the concrete and the ground panels have been bolted in place.  The first rows are taking shape and beginning to transition from horizontal to vertical. 

Written by sjet

April 28, 2008 at 1:50 am

Construction Images Day 8_4:25:08

leave a comment »

Images from Phase 2: (12) 4’x8′ sheets CNC fabricated by Continental Signs and Jared Laucks.  (www.continentalsigns.net)

2 Full strips (27 panels) have been folded, riveted and bolted together.  5 strips still to be assembled.  There are 187 total panels in the canopy structure.

 

Written by sjet

April 26, 2008 at 12:31 am

Posted in Uncategorized

DWGS_4:23:08

leave a comment »

 

 

Written by sjet

April 23, 2008 at 10:17 pm

Posted in Documentation

Construction Images_Day 6_4:23:08

leave a comment »

Images from the ground work construction.  2″ Concrete base to stabilize ( 8 ) 5/8″ x 2′ threaded rods driven into grade.  Leveling nuts and gaskets will secure the base panels to the threaded rod. 

 

Written by sjet

April 23, 2008 at 10:01 pm

Posted in Construction

Construction Images_Day 2_4:18:08

leave a comment »

5 rows down 22 to go!  Avergaing roughly 2-3 rows per hour. 

All bolts still need to be tightened to fix curvature in place.

 

Written by sjet

April 18, 2008 at 8:32 pm

Posted in Construction

Construction Images_Day 1_4:17:08

leave a comment »

A few images from the final assembly.  5 sheets have been CNC routed and delivered, 20 still out for fabrication.

Many thanks to the material sponsor, Alliance Metals (www.alliancemetals.com), for providing ( 28 ) 4′ x 8′ sheets of 1/16″ white aluminum!

Written by sjet

April 18, 2008 at 3:52 am

Fenestration Studies_4:1:08

leave a comment »

A number of fenestration studies.  Aperture dilation based on views and optimal sunlight.

7.jpg

6.jpg

1.jpg

2.jpg

4.jpg

circ_1.jpg

Written by sjet

April 1, 2008 at 5:30 am

Mock-up 4_3:29:08

leave a comment »

Images from a fourth full scale mock-up from 1/16″ aluminum sheet.  (1) 4’x8′ sheet including 9 full panels and folded connections.

Many thanks to Jared Laucks and Continental Signs for fabrication.

dscn5831_small.jpg

dscn5836_small.jpg

dscn5839_small.jpg

dscn5856_small.jpg

dscn5878_small.jpg

dscn5882_small.jpg

dscn5889_small.jpg

dscn5893_small.jpg

dscn5895_small.jpg

dscn5898_small.jpg

dscn5902_small.jpg

dscn5916_small.jpg

Written by sjet

March 29, 2008 at 7:55 pm

Posted in Uncategorized

Small Scale Model_3:26:08

leave a comment »

A few images from a small scale paper strip model looking at overall weight distribution & surface balance.

dscn5750_small.jpg dscn5757_small.jpg dscn5755_small.jpg

Written by sjet

March 26, 2008 at 8:46 pm

Connection Update_3:25:08

leave a comment »

newconnection_2-copy.jpg

newconnection_1-copy.jpg

Written by sjet

March 25, 2008 at 7:04 pm

Posted in Connection

Fenestration_Update 3:24:08

with one comment

 fenestration-board3.jpg

Versions of fillet dilation based on occupant view and optimal sunlight. 

fenestration-board2.jpg

Possibilities for fenestration aperture.

fenestration_gradient.jpg

Possible panel fillet dilation based on occupant view and optimal sunlight.

Written by sjet

March 25, 2008 at 4:53 am

Mock-up 4 Preperation

leave a comment »

A nested sheet almost ready for fabrication, still without the proper fenestration.  (1) 4′ x 8′ aluminum sheet @ 1/16″ thickness.

sheetpdf2.jpg

-Nesting obviously becomes a major restriction, a possible direction for developing the connection efficiency!

Unrolled panels & folding connection…

nest.jpg

Written by sjet

March 20, 2008 at 11:31 pm

Connection Update_Mock-up 3

leave a comment »

A few images from a third full scale mock-up, looking at 1/16″ aluminum’s folding capability.

Many thanks to Jared Laucks & Continental Signs for fabrication and photos!

dsc03736_small.jpgdsc03737_small.jpg

dsc03747_small.jpgdsc03753_small.jpg

Written by sjet

March 20, 2008 at 10:26 pm

Posted in Uncategorized

Full Scale Mock-up v.2

with one comment

A few images from the second full scale mock-up…

img_0475_small.jpg

img_0483_small.jpg

img_0467_small.jpg

Conclusions:

-1/8″ Aluminum may be excessive in weight for panel + tab connection

-A perforated score @ 1/8″ aluminum is very difficult to bend.  Either the panel must be 1/16″ thickness or use a channel notch, which may only allow for 90 degree folds.

-Folded connection is very clumsy looking, may go for a triangular shape as well as cut away most of the back face of the connection b/c the first fold restrict almost all of the force

Written by sjet

March 11, 2008 at 5:26 am

Tab Connection_2

leave a comment »

A new tab connection which folds back on itself to restrict unfolding. A full scale mock-up has been sent for fabrication.

connection_fold1.jpg

mockup2.jpg

Written by sjet

March 9, 2008 at 2:02 am

Posted in Connection

FEA Analysis_Prelimary Attempt 1

with one comment

A few images showing very preliminary FEA analysis.

axial-stress2.jpg

Axial stress with an aluminum pipe cross section.

shear-stress2.jpg

Shear stress with an aluminum pipe cross section.

shear-stress3.jpg

Axial stress with an aluminum “T” cross section.

axial-stress3.jpg

Shear stress with an aluminum “T” cross section.

Analysis includes:

-Evenly distributed loading (Inaccurate!)

-Axial & shear loading calculation for 2 cross sections

-“T” cross section appears to resist more loading and force high stress areas to the lower right landing of the surface.

-Provides generic feedback for the surface generation….

Written by sjet

March 4, 2008 at 9:07 pm

1/8 Scale Assembly

leave a comment »

More images from the 1/8 scale model….

dscn5736-copy.jpg

dscn5740-copy.jpg

dscn5735-copy.jpg

Conclusions:

-The tab connection does not restrict unfolding and requires further bolting to ensure stability.

-A tab connection can reduce overall elements from 1000+/- to 250+/- as connections are incorporated in the panel.  This probably doesn’t save money in material cost but drastically reduces assembly time and frustration!!

-Tabs were scaled based on panel edge lengths and therefore varied the size & hole distances.

-The fenestration seems to weaken the panel if it occupies a large percentage and nears the edges.

-Material width was not accounted for and would result in a 9″ increase in the overall full scale length!!  This means that all panels must be reduced by 1/8″ to account for the 1/8″ tab fold, or the fold must be set inward.

Written by sjet

March 3, 2008 at 2:42 am

Posted in 1/8" Model

1/8 Scale Fabrication

leave a comment »

An update with a few images of a full day of lasercutting the 1/8 scale model…

                  dscn5731-copy.jpg

dscn5723-copy.jpg

dscn5721-copy.jpg

dscn5699-copy.jpg

                 dscn5689-copy.jpg

                 dscn5716-copy.jpg

Written by sjet

March 1, 2008 at 1:58 am

1/8″ = 1′ – 0″ Scale Model Preparation_2:26:08

leave a comment »

 8th1.jpg

8th2.jpg

The code has been updated to:

-nest all text & panels

-Control the orientation of the connection, either inside or outside

The 1/8″ scale model will use a “tab” connection as the scale is too small to demonstrate the accurate connection.  The tab connection also utilizes the possibility of folding each panel with 1/8″ scale materials.  Two holes are provided in the tab connection to account for a standard staple to replace the full scale bolts.  The tab restricts almost all axis of rotation and should provide an estimate in the overall structural efficiency of the form.

sheet_2-copy.jpg

1 of 4 nested sheets including:

-Tab connections

-Fenestration

-Unique 2 digit number

nest8th2.jpg

Written by sjet

February 27, 2008 at 1:41 am

Connection Nest_2:25:08

leave a comment »

 srf_grow5-copy.jpg

srf_grow6-copy.jpg

The code has been updated to:

-add flat panels

-add connections on all interior panels, all edge panels still to be updated

-add three-digit annotation

-flatten and array for nesting on sheets

Still to be updated:

-edge connections

-flatten and array text

-add connection holes & tab

-possibly align perpendicular to panel

Three versions of the connection through code….

connection_code_1.jpg

The most recent connection which eliminates the “U” shape and minimizes the arm lengths.  Still 2 up & 2 down with annotation.

connection_code_2.jpg

A “Z” shaped connection which may be likely to unfold….

connection_code_3.jpg

An early “T” shaped connection….

nest1.jpg

A first nesting attempt…

Written by sjet

February 26, 2008 at 3:15 am

Posted in Connection

Connection Annotation_2:25:08

leave a comment »

A few images of the connection annotation.  A three-digit number is applied to each connection describing:

-The number of the panel in the x direction

-The number of the panel in the y direction

-The edge of the panel : 0,1,2,3

annotation_1.jpg

annotation_2.jpg

annotation_3.jpg

This image shows a version with 2 connections up & 2 connections down for each panel.

A second iteration was produced which gradients the normal of the connection based on the surface curvature.  Not sure about its structural consequences but could be a nice detail….

Written by sjet

February 26, 2008 at 2:46 am

Posted in Connection

Connection Tolerance per Iteration_ 2:25:08

with one comment

conn_tolerance_1.jpg

Connection tolerance with a 10×35 panel layout.  Red dots indicating the panels connecting with a .001 tolerance. 

conn_tolerance_2b.jpg

Connection tolerance with a 10×30 panel layout. 

The tolerance between panels does not directly influence any structural or connection factor as the connections account for each condition individually.  However, this did influence the processing of coding and resulted in an interesting discovery relating to the angle and intersection of panels with relation to the various degrees of surface approximation.

Written by sjet

February 26, 2008 at 1:49 am

Posted in Connection

Surface Generation 2:22:08

leave a comment »

srf_grow3-copy.jpgsrf_grow2-copy.jpg

Two images of the progress of surface generation.  The surface now provides occupation and user interaction on a variety of levels while responding to site conditions.  The fenestration and connections still need to be resolved.

Demonstrates a 20×50 panel iteration.

Written by sjet

February 22, 2008 at 7:04 am

Surface Generation_Fitness Criteria 02:22:08

with 3 comments

A number of panel variations leading to fitness criteria based on budget, aesthetic surface approximation, structure, and overall weight.

Each iteration demonstrates a decreasing number of panels to approximate the original surface from 30×70 panels to 10×40.

Each iteration is further analyzed based on total material used and the resultant budget for material and connections.

3070.jpg

2070.jpg

2060.jpg

2050.jpg

2040.jpg

1040.jpg

Written by sjet

February 22, 2008 at 6:50 am

Connection Update_2:19:08

leave a comment »

                                        _-001_reduced.jpg

                                        _-008_reduced.jpg

The following shows six full scale 1/8″ metal connections that were CNC routed and assembled.  Many thanks to Jared Laucks for fabrication.  Two modules were tested: 1. a tooth & notch connection, 2. a bolted tab connection.

Results:

-The tooth restricts the connection from sliding into the panel notch.  This results in weakening the metal connection as it requires a bit of forcing!

-The tooth must be longer than the current 1/8″ to ensure that it does not slip out of the panel.  Though, this would make the connection even more difficult to slide into the panel notch??

-The tab for the bolt connection on the panel cannot be folded by hand because there is minimal leverage.  This requires more surface area in the metal piece OR a thinner gauge metal!!  Both will be tested for structural integrity.

-Material thickness needs to be taken into account i.e. a  slot must be cut in the panel to offset the material inside the panel and allow each panel to meet perfectly.  The metal tab for bolting also did not take material thickness into account.

-A “v” notch should be used, rather than a channel, to score the fold lines.  This would ensure that all angles meet @ 90 and minimize error.

-If a tooth & notch connection is used, a larger bottom piece should be added to ensure strength.

-Add 1/8″ to all notch cuts in the panel and metal connection as a router cannot cut square notches…this resulted in 1/8″ on both sides being unusable. 

All slight failures aside, the connection worked very well and added up to the exact angle for each panel in both axis!!!!

More to come!

                                       _-017_reduced.jpg

                                        _-032_reduced.jpg

                                        _-054_reduced.jpg

Written by sjet

February 19, 2008 at 10:01 pm

Posted in Uncategorized

Quad Mesh

leave a comment »

Tesselion uses a method of planar quadrilateral meshes to tessellate complex surfaces.  The following is a diagram showing the stages of panelization.

steps_reduced.jpg

The following image demonstrates the range of possibilities using the Tesselion method of panelization.  The middle range shows the target surface generation for construction in mid May.  The surface criteria fall between limitations in timeline, budget, and small scale structural concerns, otherwise practically any surface can be constructed using the Tesselion method. 

diagram_reduced.jpg

Written by sjet

February 17, 2008 at 6:40 pm

Fenestration Studies

with one comment

Fenestration Studies showing aperture dilation and gradient patterns of optimal environmental conditions.

gaussian_1_diagram.jpg

1-copy.jpg

2-copy.jpg

3-copy.jpg

Written by sjet

February 15, 2008 at 4:21 am

Posted in Fenestration

Material Investigation

leave a comment »

material_fiberglass.jpgmaterial_acrylic.jpg

material_luan.jpg

material_poly.jpg

Material Investigations looking at three criteria:

-Weight/Density

-Cost

-Strength

Written by sjet

February 14, 2008 at 5:50 am

Connection Tests_2:12:08

leave a comment »

connection_1b.jpg

connection_2b.jpg

connection3b.jpg

connection_4b.jpg

connection_5b.jpg

A number of connection options to solve the following criteria:

-Resist two axis of rotation

-Resist panels from pulling away from each other

-Maintain structural integrity as it may carry most of the load

-Fabrication with either 3 axis CNC router or laser cutter, requiring flat sheet material

Written by sjet

February 13, 2008 at 4:41 am

Posted in Connection