06. 3D Scanning & Printing

Tasks

Group assignment: - test the design rules for your 3D printer(s) Individual assignment - design, document, and 3D print an object that could not be made subtractively (small, few cm3, limited by printer time) - 3D scan an object Extra credit - print scanned object

Outline

Scanning

Physical touch probe Binocular vision - 2 eyes or 1 eye + movement Point correspondence - structured light, laser + camera

Structured light, laser, LiDAR & long distance scanning

Printing

FFF - Fused Filament Fabrication

Alternatively known as the trademarked term FDM (Fused Deposition Modelling), this is an additive processes that deposits layers of melted thermoplastic to build a 3-dimensional object. AS the material filament is heated to its melting point it is extruded out of the nozzle onto the build platform in accordance with the reference design. Using a 3D CAD model, exported as a STL or 3mf file, the model is inserted into a slicing software, which communicates build instructions with the printer using gcode. The slicing software generates the layers/paths that the printer will follow to build the desired shape. By further defining the printer settings print quality, finish, strength etc can be altered, additionally more complex designs may use functions like supports or brims to ensure a successful print. After the build is complete additional support material is removed and the print can undergo finishing processes such as using solvent, sanding or electrostatic paint.

Print Considerations

Material / temp / speed

Hight of 1st/Layer - Reducing layer hight can increase resolution/finer details, but increases how many layers the model is split into, therefore increasing print time. The 1st layer should be thicker (double) the following layers to ensure the print adheres to the build plate

Layer width & precision

Solid layers - Reinforces top & bottom surface by ‘colouring in’ the entire layer, infill pattern isnt used. Can be set to distributed throughout model to increases strength by increasing bonding between layers

Bed levelling - Distance of nozzle to bed (z-axis) is consistent = good adhesion, more consistant/successful prints. Can be tested using paper grab test

Base - skirt (offset outline that clears initial extrusion from nozzle), brim (attached outline to increase adhesion surface area, raft (depositis a platform the model is built upon due to small surface area or complex geometry)

Overhangs - degree at which material can overhang its base, whilst maintaining quality & not using supports

Bridging - deposition of material across 2 bases of support, crossing span quickly or upping fan cooling can help prevent sagging

Clearance - minimum distance between parts whilst maintaining freedom of movement

Supports - provide additional structures e.g. trees or pillars to facilitate build during printing, removed afterwards. Adjusting settings e.g. z distace from model, can imporve surface finish/ease of removal.

Infill Shape - Shape used to reinforce internal structure, can be changed to effect print time & strength e.g. hexagons or triangles . freeflow allows 3 dimensional movement of a fluid within an object Infill Overlap -

Shells - Number of walls the print has before infill area Shell Overlap -

Surface finish - resultant finish of print due to applied settings e.g. ironing uses a final pass to melt the distinction between layers

Vase -

Design Considerations

Dimensional accuracy - Internal dimensions get smaller

Anisotropy - Consideration of how a objects mechanical properties can be impacted by layer adhesion, print orientation & microstructural variation. The bond within a layer is typically stronger that the bond between layers resulting in weakness in the Z direction. in the XY plane tensile strength can be up to 50% higher, with mechanical loads applied parrallel to layers performing better than loads applied perpendicular(along z-axis)

Forcing internal structures - By adding irregular cylinder rods into the structure of a CAD model the resultant part can have higher strength. The rods act like fibers across the model & the slicing software reinforcing the models defined shapes

Minkowski Sums ?

Pause & place - Used to add magnets, other parts or fixings into a model during the print so that it is enclosed within the finished print

Print in Place - building mechanism into model that functions straight from the build plate

Multiple filament printing - using different colours or different materials within one print e.g. soluable supports that can be removed without affecting the main model

Materials

PLA, ABS, TPU

Vat Polymerization / Resin Printing

Group of 3D printing processes that uses a light source to cure photopolymer resin in layers corresponding with the digital model. Once printing is complete the object must be cleaned and post-cured to ‘lock’ the material properties

SLA - Steriolithography

The original 3D printing, they use mirrors to rapidly aim laser beams & solidify a cross section of the model bit by bit. Whilst a slower process it is still commonly used, particularly in applications that require a larger format printer or higher grade resin. The laser can cure ‘stronger’ resins and reach cross larger distances without losing power than the following processes

DLP - Digital Light Processing

Uses a digital light projector to flash an image of the layer & cure it all in one go. generally faster than SLA its used for producing higher volumes of parts within a batch

MSLA - Masked Steriolithography

Similar to DLP but uses a LCD screen with an array of emitters. Photomask is composed of sqaure pixels which determin the granularity of the print

Other Printers

direct energy deposition (DED) wire arc additive manufacturing (WAAM) Selective laser sintering (SLS) Binder Jetting bioprinting, clayprinting

Class

Printer Testing

Printer Spec:
- Flashforge adventurer 4
- Nozzel diameter = 0.4mm
- Bed size =
- Magnetic build plate
- Slicer = Flashprint 
- Material = PLA

Benchy boat

Benchys are a quick print standard model consisting of design elements that ‘challenge’ the 3D printer, by comparing the outputs of varied base parameters e.g. temperature, adhesion, bed levelling, these settings can be calibrated. I downloaded it as an 3mf before slicing it with the suggested properties. By printing 5 boats & varying the slicer settings I am able to compare the effect on the resultant print. Throughout these tests I kept consistent the number of walls (shells) and infill shape/percentage (hex 10%) benchy_download

Benchy 1

This boat was printed useing the settings suggested on the above linked site. There is some filament stringing between the print, the ‘number plate’ is to low resolution to read and the base text is legible but flat & unclear. Bridging was good.

Printer Settings: 1
- Time = 1h 19min
- Nozzle temp = 220C
- Bed temp = 55C
- Layer hight = 0.2, Layer 1 hight = 0.3
- Layer width = 0.4, Precision = 0.1
- Walls = 2, Overlap = 30%
- Top/btm layers = 3/2
- Infill = 10% hex, Overlap = 15%
- Print speed = 50mm/s
- Travel speed = 100mm/s
- Bridge speed = 50%
- Fan = ON 25%

Benchy 2

layer width was halfed to 0.2, 1st layer hight remains consistant at 0.3, all other layers are decreased by 0.05 to 0.15. Compaired to benchy 1, benchy 2 was printed at a faster pace to componsate for the finer layers, print speed doubled with travel & bridge speed increasing by 50%.

The resultant print is weak, the chimney broke off during removal from the build plate & the surface is pockmarked where the shells to thin. Along with the wispy stringing that remains from benchy 1 this suggests the nozzle temp is to high

Printer Settings: 2
- Time =
- Nozzle temp = 220C
- Bed temp = 55C
- Layer hight = 0.15, Layer 1 hight = 0.3
- Layer width = 0.2, Precision = 0.1
- Walls = 2, Overlap = 30%
- Top/btm layers = 3/2
- Infill = 10% hex, Overlap = 15%
- Print speed = 100mm/s
- Travel speed = 150mm/s
- Bridge speed = 75%
- Fan = ON 25%

Benchy 3

Maintaining layer hight & width from the previous print (becnhy 2), Benchy 3s speed was decreased to the initial speeds used in test 1. Nozzle tempuratre was decreased from 220C to 215C to address the issues of webbing in the previous prints

Printer Settings: 3
- Time =
- Nozzle temp = 215C
- Bed temp = 55C
- Layer hight = 0.15, Layer 1 hight = 0.3
- Layer width = 0.2, Precision = 0.1
- Walls = 2, Overlap = 30%
- Top/btm layers = 3/2
- Infill = 10% hex, Overlap = 15%
- Print speed = 50mm/s
- Travel speed = 100mm/s
- Bridge speed = 50%
- Fan = ON 25%

Benchy 4

To increase model detail whilst maintaining structural integraty. Found suggestions that layer width & precision XXXX. Both 1st and resultant layers are decreased, increase the number of solid layers on the top & bottom to counteract the lower layer hight being used. Bridge speed decreased by 10% to see effect on sagging. Option to reinforce small areas < mm2

Printer Settings: 4
- Time =
- Nozzle temp = 215C
- Bed temp = 55C
- Layer hight = 0.1, Layer 1 hight = 0.2
- Layer width = 0.4, Precision = 0.3
- Walls = 2, Overlap = 30%
- Top/btm layers = 6/5
- Infill = 10% hex, Overlap = 15%
- Print speed = 50mm/s
- Travel speed = 100mm/s
- Bridge speed = 40%
- Fan = ON 25%
*reinforce small areas

Benchy 5

Wall overlap decreased by 10%, infill overlap increased by 15% (double). Bridge speed increased by 10% to see effect on sagging. Option to reinforce small areas < mm2 and number of solid top/bottom layers is maintained

Printer Settings: 5
- Time =
- Nozzle temp = 215C
- Bed temp = 55C
- Layer hight = 0.1, Layer 1 hight = 0.2
- Layer width = 0.4, Precision = 0.3
- Walls = 2, Overlap = 20%
- Top/btm layers = 6/5
- Infill = 10% hex, Overlap = 30%
- Print speed = 50mm/s
- Travel speed = 100mm/s
- Bridge speed = 60%
- Fan = ON 25%
*reinforce small areas

Benchy Ranking

Speed Finish Layer visability Chimney Number plate Base text Bridging Portholes Detailing

Clearance

By testing the distance at which 2 parts lose there ability to move freely the clearence distance can be determined. Considering this in the design increases likelihood of successful print in place mechanisum. The tree supports worked more effectively when the model was arranged in the slicer on its back, as oppose to up right on its base. In this configuration the supports base has contact with the build plate, making it a more effective support structure whilst reducing post print processing. Supports where applied using slicer auto mode & excess removed manually e.g. supports within numbers.

The supports used for the column, between the rod & base, where a challenge to remove compared to all other supports coming off with minimal effort. Pliers where required for removal and a rough surface finish was left. I suggest this is caused by to many supports to close together, allowing bonding between supports which increase there adhesion to the model.

From this test I found that the minimum clearence distance between 2 parts without friction is 0.4mm. 0.3 = slides well but grabby on rotation 0.2 = resistant to roation 0.1 = no movement

Assignment

Implementation

Manipulating a trees ‘face’ to create a door knocker. I found this object nailed to a tree, I chose it due to the interesting texture and resemblance to a ornate door knocker

Scanning

I scanned the object using the app surescan and exporting in .ply format. Whilst I tried several apps, I found all of them to restrict available file types by payment tier & difficult to ascertain when a good level of detail was achieved

Point Cloud - blender

• Extract points
• gave up & switched to rhino / meshes

Mesh - Rhino

Version 1

• insert as
• Mesh split
• Surface from 4 points

  Patch - created UV map (didn’t use)

• Extract mesh edge/change layer /make2d
• booleen intersect

• selnaked edge

• outline mesh - extrude crv - srf output

• match mesh edge - join
• shrink wrap (make water tight)

print 1

Used to ascertain detail level from print settings & mesh complexity, after this print the shrink wrap settings were adjusted to allow for more deatailing

Version 2

Redid shrink wrap from the original mesh & stretched the shape across its width to change the proportions, allowing the ring (detailed below) to be proportional with the face. Also added a back plate for the face to sit on with mail attachment slots - taurus - sephere - mesh - mesh offset - mesh Boolean difference - Base extrusion from offset edge - cplane - sketch - extrusion - intersect - delete/join

print 2 Happy with the sizing & quality of this print, but there wasn’t enough clearence between the ring & nose, so the part could not move.

Version 3

Separate ring from face drew screw slots intersecteded with base plate mesh to form hollow with partial overhang

print 3 ring printed seperatly 1st, followed by face which was pause at est layer 90 to insert ring. This didn’t work as while the ring fit, the nozzle head didnt have enough space to complete the print around/over the ring. The print was canceled at this point.

Version 4

Before printing the 4th version I did a clearence benchmark test to ascertain distance required between nose & ring to get a freemoving part, this was postioned hisher on the zaxis to reduce area of interference with top lip when printing. Additionally I intersected the face mesh to contain cylinders to hold magnets, to be inserted mid print, in the tounge & ring knocker.

New ring revolved ring, adjusted placement working in section view (Section View = clipping plane - mesh split) mesh boo diff with face mesh then mesh split to remove pipe face

Magnet hole 1 - extruded cyclinder with base, meshed - mesh Boolean split - dup edge mesh split with curve Magnet hole 2 - extrude cylinder - cap - mesh Boolean split - deleted

Key Terms

Slicer

Software used to split CAD model into thin horizontal layers for printing

GCode

Format used to communicate with printer

Anisotropy

the directional dependance of a materials mechanical properties, how a material behaves differently depending on its orientation to a applied load.

Isotropic

Materials that have uniform strength and stiffness regardless of directions

Thermoplastic :

Thermoset Plastic :

Photopolymer

Plastic/polymers that react to light

Voronoi diagram voronoi_generator

a

Minkowski Sums

? make things fit by adding shape

---

Notes