Embedded Files
Nature|Replication
Nature|Replication
Spring 2014 MAE 156B Sponsored Project
Jacobs School of Engineering
University of California, San Diego
Sponsored by: Melinda Barnadas of Magpie Studio
A Special Thanks: Nathan Wade of SME Robotic CNC Machining and 3D Printing LAB
Patrick Wilcox of Swinerton Builders Construction
Chao Han, PhD of VRMesh
Background
Magpie Studio has been producing plant and animal models for natural history museums for over ten years. Natural history museums have two missions: to house specimen collections used for biological research and to educate the public through exhibition display. A key element of exhibition display is the construction and presentation of accurate models. Artisans trained in sculpting and mold-making, along with biologists have historically made the models. Digital archiving and rapid reproduction methods will re-invent both museum exhibitions and taxonomic research.
Magpies current project includes the creation of full scale models of two of the world’s largest and rarest botanical blooms – Amorphophallus titanum or titan arum and Rafflesia arnoldii. These models are being developed to juxtapose and highlight the historic glass flower exhibit at the Harvard Museum of Natural History.
The R. arnoldii can reach a diameter of 1m (3ft). Its bloom can take several months to develop, and last only 48 hours. The titan arum may grow to be 3m (10ft) tall. It’s A sexual, however its flowers never bloom simultaneously and must be cross-pollinated from adjacent blooms. It can take up to seven years for the flower to bloom, also lasting only a few short days. Due to the difficulty these processes, both plants are becoming endangered.
Objectives
Both flowers are to bloom within the year, which may not happen again for some time. We are tasked with developing a reliable and repeatable method for modeling the Amorphophallustitanum that can be applied to both flowers while problem solving and trouble-shooting the process prior to their bloom to avoid any unforeseen issues. The main objective of the project was to create a reliable replication process that can be easily followed. Also to help exhibit a need for further development and funding in different arts by experimenting with 3D scanning, rapid prototype computer numerical control (CNC) milling, material exploration, computer generated mold development, 3D scan repair software, and constructing a full-scale model.
Requirements
Requirements
Develop a reliable and repeatable method for replicating both the A. titanum and R. arnoldii flowers, while problem solving and trouble-shooting the process beforehand to avoid unforeseen issues during the actual modeling process. As part of this problem solving and trouble-shooting process the following requirements were derived:
Create rough models of both flowers
Scan rough models
Develop and recommend “scanning” or 3D digital image production equipment
Edit and cleanup resultant 3D images using software
Section edited 3D images using software so as to build models in parts
Develop negative molds of key characteristic and problematic plant features using software
Recommend and develop software work flow
Export edited files into a CAM software where the Titan Arum will be reproduced by milling foam using a KUKA 6-axis CNC mill.
Maintain an accuracy of end milled models to within 3.175mm (0.125”)
Deliverable
The purpose of this project is to help the project sponsor in developing a grant for further research and development to complete this process. To facilitate this endeavor we are to provide the following:
Develop a method that can be handed on and followed for generating a 3D image of these plants as well as a full scale model.
Experiment and create files, models, and samples that may be used as incentive and to aid in project exhibit.
Final Design
The final design focused on creating a reliable and accurate process for the replication of the Amorphophallus titanum, that can later be adapted for the Rafflesia Arnoldii. The Amorphophallus titanum was separated into three main areas of focus, the spadix, spathe and male/female flowers, which can be seen in figure 1 below.
Figure 1: Names of Each Section of the Flowers
Images sources: http://botanicus.org/page/441733 (main flower); http://www.thegorgeousdaily.com/amorphophallus-titanum/ (Male/Female
The final design included a reliable and accurate method for generating a 3D model of the Amorphophallus titanum, using multiple 3D scanners and Structure from Motion (SfM) technology, file manipulation of the models, and replication by CNC milling or 3D printing, as shown in figure 2 below.
Figure 2: Flowchart of Replication Method
Once the computer generated 3D models were created, they were cleaned and edited, using various file manipulation techniques. The manufacturing method for each section of the three sections for the Amorphophallus titanum were then determined. The three different techniques were explored and used during the recreation process. The male/female flowers were recreated by the 3D printing method, while the spadix and spathe were recreated by CNC milling, using the KUKA 6-axis robotic CNC milling from Robotic Solutions Inc. The spadix was CNC milled using the first technique, of milling a positive mold. The spathe was created using the second CNC milling technique by creating three part mold, using negative molds.
Results
The spadix was milled as a positive mold from the 3D model of the spadix obtained from the 3D scanner and SfM software. Low density 2 lb. EPS foam was used for the spadix and milled into 3 sections which were glued together. The spadix was scaled to a height of 1.5748 m (62 in), to replicate the actual size. The figure below shows the final result of the milled spadix.
Figure 3: Final Milled Spadix
Due to the thin structure the spathe, it was ill-suited for being milled as a positive. As such an alternative method of production was to create a mold to be used for casting or layups. Due to the intricacy of the edge of the image it was difficult to create a common plane for which to split the mold. A three part mold was establish that was well suited for structure of the spathe, and was created using the following process:
•Import cleaned and repaired image into Maya 3D animation software
•Delete polygons around rim of spathe along the interface of the inside and outside of image
•Separate image
•Invert and extrude images to create mold using Meshlab
Figure 8: Final Case of Spathe
3D Printing Method
The male and female flowers presented a unique problem, because despite all of the 3D capture methods visited, none were capable of accurately capturing these smaller details, as can be seen in the 3D image capturing table. Fortunately as seen in figure 9, there is a naturalistic pattern in both the male and female flowers.
Figure 9: Naturalistic Pattersns of Male/Female Flowers
As such, small sections of the flowers can be scanned and imported into a 3D imaging software and replicated producing a digitally created pattern mimicking the naturally created one. The resultant can then be printed.The male/female flowers were constructed using Solidworks and then 3D printed. Figure 10 below, shows the final results of the 3D printed model of the male/female flower.
Figure 10: Final 3D Printed Model of Male/Female Flowers
Below, the first video shows the scanning process using the Trimble TX5 and the second video shows the milling of a scaled version of the scanned spadix using the KUKA 6-axis robotic CNC mill.
Page updated
Report abuse