Mechanical Simulation and Optimisation of a Biomedical Implant using SIMUFACT ADDITIVE Software
This project was done and submitted to the Faculty of Science, Engineering & Technology in partial fulfillment of the requirements for the unit ADM80020 Additive Manufacturing and Tooling of the Master of Engineering Science (Advanced Manufacturing Technology) postgraduate course.
This will guide someone to easily replicate this particular study.
For more details, send me a DM.
I was tasked to do a Mechanical Simulation and Optimisation for a given Biomedical Implant model.
This was a Metal Powder Bed Fusion Simulation for the said model using 海克斯康 's Simufact Additive Software.
The following is a discussion in detail of the mechanical simulation setup and optimisations that were used.
As an overview, I did two processes for the mechanical simulation:
Process 1: Initial Simulation Setup
Process 2: Support Optimisation
Process 1: Initial Simulation Setup
The objective of this initial simulation setup is to prepare the simulation environment and initial conditions.
1. Setting up Process Properties
> Simulation Configuration: Mechanical
> Type of Simulation: Manufacturing (Accurate)
> Manufacturing Process Stages, in order: Build > Cutting > Support Removal
In this simulation workflow, I first addressed the setup of process properties. I then configured the simulation type as mechanical, specifically focusing on manufacturing with an emphasis on accuracy. Within this manufacturing simulation, I defined a sequence of process stages, each executed in a specific order:
This systematic approach to process property setup allowed me to comprehensively model and evaluate the mechanical aspects of the manufacturing process, ensuring the accurate representation of each stage within my simulations.
2. Setting up Machine Properties
I then configured the machine settings, including printer specifications and parameters relevant to the 3DSystems ProX200 printer, as follows:
3. Setting up Machine Cost
I then defined the hourly machine cost and fixed mixed cost per part for cost estimation.
These cost approximations were as close as possible to market values as of Q3 2023.
4. Importing, Scaling, and Positioning of CAD Model
To start the simulation, I first brought in the 3D CAD model of the biomedical implant. I then made sure the model was the right size and positioned correctly for the printing process.
This step is essential to ensure everything fits as it should during simulation and real-world production.
This involved two steps:
I began by setting the units to millimeters (mm) for consistency. Next, I imported the part, keeping all default settings as is. Following that, I applied a scaling factor of 0.70 and repositioned the model so that it approximately resides at the center of the printer area and is well-contained within the printer's boundaries.
This adjustment ensured that the model was appropriately situated for the simulation within the printer's working space.
> Units: mm
> Import part, remain everything default
> Scale: 0.70
Positioning part at the center
From the default position, translate the part accordingly:
5. Choosing Material
From the Simufact Additive Material library, the material chosen for the part (Biomedical Implant Model) is TiAl6V4.
6. Setting up Material Cost
We then specified the material cost based on the chosen material as follows:
Again, this cost approximation was as close as possible to market values as of Q3 2023.
7. Support Generation
During the simulation process, I generated support structures using the Simufact method.
The aim was to ensure successful printing by providing stability and reinforcement.
I then set the support radius to 0.23 mm to achieve this.
> Method: Simufact
> Support radius: 0.23 mm
8. Support Removal Setup
I also addressed support removal setup, focusing on defining the steps required for removing support structures after printing. I utilized a cutting method, specifically directional cutting along the x-axis.
The cut was set at a height of 1 mm, and the cutting direction was from the left to the right (+X).
This configuration ensured a methodical and effective approach to removing support structures post-printing, contributing to the overall quality and usability of our 3D-printed objects.
Cutting of the support:
9. Setting up Build Properties
I then configured the Build Properties to ensure accurate material selection and appropriate settings for the project.
I ensured that the correct assigned material was chosen from the database. Then, as illustrated in Figure 10, I opted for the default settings regarding layer parameters and inherent strains.
It's worth noting that a recommended industry practice is to perform a calibration test by initially printing simple beams and then utilising the inherent strains derived from the test results. This approach enhances the precision of our settings and contributes to the overall quality of the printed objects.
10. Meshing
In the meshing step, I generated a mesh tailored for mechanical simulations, relying on the default settings suggested by the system to optimize the accuracy of our model.
The mesh parameters we adopted were as follows:
For the mesh type, I utilised the default uniform meshing approach to ensure consistent and even mesh distribution.
To strike a balance between precision and computational efficiency, I adhered to the system's recommended coarsening level of 2.
For Voxel Size (x/y/z), I set the voxel size to the default value suggested by the system, which was 1.1772 mm for all three dimensions (x, y, and z).
This default setting allowed me to capture fine details while efficiently managing computational resources for the simulation.
Mesh Generated
I then performed volume fraction clipping to confirm the solid nature of the interior within the model.
Note that the model is positioned diagonally, as evident in its orientation as shown in Figure 3.
When I apply y-axis clipping, it selectively reveals only the portion that intersects with the clipping plane. However, as I move the clipping position across the entire model, I consistently observe a uniform red appearance throughout (see Figures 13 and 14).
This uniformity indicates that the model is entirely filled with material, and the voxel generation process results in a solid, uninterrupted structure.
11. Setting up Numerical Parameters
When configuring numerical parameters specific to my simulation, I considered the available hardware, which consisted of 8 cores.
To optimize computational efficiency, I strategically divided the model into 4 distinct domains.
领英推荐
This division allowed me to fully utilize all 8 available cores by assigning 2 cores per domain.
By doing so, I effectively harnessed parallel processing capabilities, enhancing the overall performance of the simulation. See Figure 15.
12. Doing a Model Check
Before moving forward, I conducted a model check to make sure there were no mistakes in the CAD model or simulation setup.
This step was crucial for ensuring that everything was error-free and that my simulations would be accurate and reliable. See Figure 16.
Simulation Report: Process 1 - Unoptimised
Below is a detailed overview of the simulation in Process 1, where I employed default parameters suggested by Simufact Additive.
Assumptions
Key Results
Total Build Time: The total build time for Process 1 was 20,307.40 seconds (approximately 5 hours and 38 minutes).
This parameter is vital for evaluating the efficiency of the printing process. The extended build time suggests that optimising certain parameters may reduce production time.
Total Costs: The total cost of production amounted to $768.82. This figure encompasses various cost factors, such as machine operation and material expenses.
This serves as a crucial metric for assessing the economic viability of the 3D-printed object.
Shape Comparison: I conducted a comprehensive analysis of the model's shape to assess its accuracy. Surface deviation ranged from a minimum of -1.24 mm to a maximum of 0.82 mm.
These values represent the variation between the printed model and the intended design. Minimising surface deviation is essential for achieving precision in 3D printing.
Total Distortion-Displacement: This parameter gauges how much the printed model deviates from its intended position. The maximum displacement observed was 4.43 mm, while the minimum was 0.61 mm.
Reducing these displacements is vital to ensuring that the printed object aligns with the design specifications.
AM-Layer Z Displacement: With a maximum displacement of 0.08 mm and a minimum of -0.05 mm, this parameter provides insights into the layer-to-layer alignment of the 3D-printed model.
Minimising layer displacement is essential for achieving structural integrity.
Equivalent Stress: I analysed the stress distribution within the printed model. The maximum equivalent stress recorded was 1,307.53 MPa, while the minimum was 0.00 MPa.
Understanding stress distribution is crucial for assessing the mechanical performance of the object.
Total Supports Volume: This metric quantifies the volume occupied by the support structures, which assist in the printing process. In this process, the total support volume was 3,286.20 mm3.
Optimising support structures can potentially reduce material usage and printing time.
Process 2: Support Optimization
In Process 2, my primary goal was to enhance the quality of the 3D printing process by refining and optimising the support structures that were initially generated in Process 1.
This optimisation step aims to improve printing precision and overall object quality.
1. Copying Process 1 Without Results
In the initial step of Process 2, I duplicated the simulation setup from Process 1 for consistency and reference.
This approach allowed me to maintain continuity in my simulation workflow and build upon the foundation established in the previous process.
2. Doing Support Optimisation
In the next phase, I optimised the support structures to further improve the 3D printing process.
To accomplish this, I utilised the "Support optimization" feature in Simufact Additive under the "Manufacturing and Optimization" category.
This method allowed me to fine-tune the support parameters for enhanced printing quality.
Specifically, I set the minimum support radius to 0.1 mm and the minimum volume fraction to 10%.
These adjustments optimised the support structures and ultimately improved printing results.
Below are the steps in Simufact:
Manufacturing and Optimization > Optimization > Support optimization
3. Meshing
Similar to the approach employed in Process 1, I utilised a uniform mesh type, with uniformity maintained throughout the model. The voxel size remained at 1.1772 mm in all dimensions (x, y, and z), consistent with the previous process.
As a result of this meshing operation, a detailed mesh was successfully generated, as depicted in the accompanying figure. This mesh consisted of 5,972 voxels, providing a thorough representation of the model. Additionally, the mesh comprised 8,005 nodes, facilitating precise calculations and simulations. In total, 34 layers were created, each contributing to the overall fidelity of our model representation.
This refined mesh, tailored to the optimized support structures, was instrumental in ensuring the accuracy and reliability of my simulation results in Process 2. See Figures 19 and 20.
Next, to confirm the solid nature of the interior within the model, I conducted volume section clipping, mirroring the approach employed in Process 1.
4. Doing a Model Check
Following the completion of support optimization and mesh generation, I performed a thorough model check once more to ensure the error-free integrity of the model and its setup, now enhanced with the optimised support structures.
While I was pleased to discover that no errors were present, I did encounter a notable warning message:
"The current minimum radius for the support optimisation leads to a low coverage of the voxel bottom area."
This warning prompted the following advice:
"Please reduce the radius for the support structures in the support optimisation dialog."
However, for this second process, I chose to proceed with the simulation analysis without implementing the advice provided.
Simulation Report: Process 2 - Support Optimization
The following are detailed results, including key parameters and explanations of any changes made during the support optimization
Simulation Parameters:
Key Results
Total Build Time: In Process 2, the total build time significantly decreased to 18,742.50 seconds (approximately 5 hours and 12 minutes) compared to Process 1.
This reduction in build time indicates an improvement in printing efficiency achieved through support optimization.
Total Costs: The total production cost for Process 2 amounted to $729.71, reflecting a slight decrease compared to Process 1.
This reduction is attributed to the shorter build time, resulting in lower machine operation costs.
Shape Comparison: A shape comparison revealed remarkable improvements in the model's accuracy. Surface deviation ranged from a minimum of -0.10 mm to a maximum of 0.02 mm.
These values indicate a significant reduction in variation between the printed model and the intended design, showcasing the effectiveness of support optimization in enhancing precision.
Total Distortion-Displacement: With a maximum displacement of 0.20 mm and a minimum of 0.00 mm, Process 2 exhibited improved stability compared to Process 1.
The reduced displacement values indicate enhanced alignment with the design specifications.
AM-Layer Z Displacement: The maximum AM-layer z displacement was 0.01 mm, with a minimum of -0.07 mm.
This indicates minimal layer-to-layer misalignment, further contributing to improved structural integrity.
Equivalent Stress: The analysis of equivalent stress showed a maximum value of 1,265.40 MPa and a minimum of 0.00 MPa.
Stress distribution remained consistent, but the overall values were slightly reduced compared to Process 1, suggesting better printing quality.
Total Support Volume: The total volume occupied by support structures increased to 3,831.16 mm3, reflecting the adjustments made during support optimization.
While this may increase material usage, the resulting improvements in print quality justify the change.
Explanation of Parameter Changes
During support optimization, we adjusted the minimum support radius, a critical parameter. This change helped reorient the part and minimize distortion. Although a warning message advised reducing the radius to enhance coverage of the voxel bottom area, we decided to continue with the simulation without implementing this advice, as we aimed to evaluate the effects of the changes on the final printed object.
Conclusion
Process 1, with its unoptimized settings based on Simufact Additive defaults, provided valuable insights into the printing process and the 3D-printed model's characteristics. Notable results include the total build time, costs, shape comparison, displacement, stress distribution, and support volume. These findings serve as a foundation for further optimisation efforts aimed at improving printing efficiency, accuracy, and cost-effectiveness. The results also help in understanding the quality and mechanical performance of the printed object, setting the stage for future refinements in the simulation process.
Process 2, focused on support optimisation, resulted in significant improvements in various key parameters. The reduced build time, lower costs, and enhanced accuracy in surface deviation and displacement measurements demonstrate the effectiveness of support optimisation in enhancing the overall quality of the 3D-printed object. The adjustments made during this process were essential in achieving these positive outcomes, affirming the importance of fine-tuning support structures in 3D printing for optimal results.
P.S.
Hope you learned something new today!
For more details, feel free to DM me, and I'm more than happy to connect!
If you want to learn more info about the software used, here's the link to 海克斯康 's Simufact Additive: https://hexagon.com/products/simufact-additive
Huge thanks to Suresh Palanisamy for being the Unit Convenor of 澳大利亚斯威本科技大学 's Additive Manufacturing and Tooling Unit, and especially to Rizwan Abdul Rahman Rashid and Muhammed Abdul Khalik for their guidance and expertise.