Purdue University Graduate School

File(s) under embargo





until file(s) become available


posted on 2023-06-18, 16:47 authored by Xiaoxu ZhongXiaoxu Zhong

Autoinjectors are pen-like devices that typically deliver drug products of 2 mL or less. They shield the needle before and after use, reducing patient anxiety from needle phobia and mitigating the risk of needlestick injuries and accidental contamination. Additionally, automatic delivery ensures more consistent needle penetration depth and injection force than manual injection methods. 

To optimize autoinjector design, this thesis presents experimentally validated computational models that describe the processes of needle insertion, drug delivery, and transport of subcutaneously administered therapeutic proteins in the body. A multi-objective optimization framework is also proposed to guide the design of autoinjectors.

This thesis focuses on spring-driven autoinjectors, the most common type of autoinjector. It begins with an overview of the interactions between the spring-driven autoinjector, tissue, and therapeutic proteins. Moving on to Chapter 2, a computational model is presented to accurately predict the kinematics of the syringe barrel and plunger during the needle insertion process.

In Chapter 3, we present a quasi-steady model for the drug delivery process, which considers the rheology of therapeutic proteins. The Carreau model is adopted to describe protein viscosity, and explicit relationships between flow rate and pressure drop in the needle are derived. Furthermore, the applicable regime for the power-law model for protein viscosity is identified.

Chapter 4 quantifies the impact of sloshing and cavitation on therapeutic proteins in the syringe. Additionally, a workflow is presented to integrate available simulation tools to predict the performance of spring-driven autoinjectors. The influence of each design parameter of spring-driven autoinjectors on their performance is also discussed. 

The spring-driven autoinjector delivers therapeutic proteins through subcutaneous administration. To gain insights into the transport process of therapeutic proteins, Chapter 5 presents a physiologically-based pharmacokinetic model that has been validated against experimental data for humans and rats. The lymph flow rate significantly affects the bioavailability of therapeutic proteins. This finding highlights the importance of studying the transport of therapeutic proteins in the lymphatic system in future research.

Chapter 6 provides a multi-objective design optimization framework for the spring-driven autoinjector. The computational model is replaced with an accurate deep neural network surrogate to improve the computational efficiency.  Using this surrogate model, we conduct a sensitivity analysis to identify essential design parameters. After that, we perform multi-objective optimization to find promising design candidates.

Chapter 7 presents a model for bubble dynamics in a protein solution. An explicit expression for the bubble dissolution rate is derived, enabling extraction of the interfacial properties of the protein-coated interface from the measured bubble radii. Moreover, analytical solutions for the response of a protein-coated bubble to an imposed acoustic pressure are derived. This work provides insight into protein-coated bubbles, which are used as vehicles to deliver drugs, as active miniature tracers to probe the rheology of soft and biological materials, or as contrast agents to enhance the ultrasound backscatter in ultrasonic imaging.

At last, in Chapter 8, we introduce a model for laser-induced cavitation that considers several key factors, such as liquid compressibility, heat transfer, and non-equilibrium evaporation and condensation. Our model's predictions for the time-course of bubble radii have been validated with experimental data. Moreover, our model reveals that the reduction of the bubble's oscillation amplitude is primarily due to a decrease in the number of vapor molecules inside the bubble, highlighting the crucial role of phase change in laser-induced cavitation bubbles.

The developed computational models and framework provide crucial insights into the development of spring-driven autoinjectors and cavitation bubbles. These studies can also enhance the efficacy and safety of the delivery of therapeutic proteins, ultimately improving patient outcomes.


Degree Type

  • Doctor of Philosophy


  • Mechanical Engineering

Campus location

  • West Lafayette

Advisor/Supervisor/Committee Chair

Arezoo M. Ardekani