Stiffness prediction methods for additively manufactured lattice structures

Thesis: S.B., Massachusetts Institute of Technology, Department of Mechanical Engineering, May, 2020 === Cataloged from the official PDF of thesis. === Includes bibliographical references (pages 87-91). === Since the initial 300 pair release of the Futurecraft 4D in April 2017, adidas has scaled its...

Full description

Bibliographic Details
Main Author: Folinus, Charlotte Méry.
Other Authors: Anette "Peko" Hosoi.
Format: Others
Language:English
Published: Massachusetts Institute of Technology 2020
Subjects:
Online Access:https://hdl.handle.net/1721.1/127918
Description
Summary:Thesis: S.B., Massachusetts Institute of Technology, Department of Mechanical Engineering, May, 2020 === Cataloged from the official PDF of thesis. === Includes bibliographical references (pages 87-91). === Since the initial 300 pair release of the Futurecraft 4D in April 2017, adidas has scaled its 4D program to mass produce additively manufactured shoe midsoles. The 4D midsoles are constructed from lattice structures, and if there is variation in the manufacturing process, the structure's material and/or geometric properties may be altered. This means midsoles may have the same geometry but different material properties and thus different stiffnesses, and they may also have the same material properties but different overall stiffness due to geometric changes. The current quality control test is slow, expensive, and does not scale well. This thesis explores two potential techniques: using ultrasonic waves to determine the lattices' acoustic properties, and weighing them to determine their mass. Pulse-echo testing data for n = 8 samples shows a statistically significant (p = 0.0398 < 0.05) increase in response time due to sample stiffness. Stiffness scaled linearly with lattice mass for both physical and simulated lattices, and mass predicted lattice stiffness with a minimum accuracy of 90% across a range of simulated manufacturing conditions. An analytical framework parameterized around a bivariate normal distribution can determine accuracy of new test methods or from additional mass-stiffness data. Lastly, cost minimization is presented for a hybrid test protocol which combines mass testing with secondary testing for rejected samples. At specification limits of ±1[sigma], the hybrid test achieves 99% accuracy at 69.8% of the cost for the current test. Increasing the specification limit to ±2[sigma] reduces cost further, achieving 99% accuracy at 16.4% of the current cost. === by Charlotte Méry Folinus. === S.B. === S.B. Massachusetts Institute of Technology, Department of Mechanical Engineering