P15441: Mini-Air
/public/

Detailed Design

 Table of Contents 1 Prototyping, Engineering Analysis, Simulation 2 Drawings, Schematics, Flow Charts, Simulations 3 Bill of Material (BOM) 4 Test Plans 5 Risk Assessment 6 Presentation

Prototyping, Engineering Analysis, Simulation

The technical validation matrix below shows the status of the prototyping, analysis, and simulations that were performed in order to validate the Engineering Requirements for this project and if we met them.

Technical Validation Matrix

Airflow Subsystem

To determine the airflow for the system, we first needed to understand the pressure drop across the heatsinks. Knowing this, we can overlay the fan performance curve with the impedance curves of the heatsinks to find our operating point and actual CFM output. To do this, the formulas located at this link were used. They calculate the pressure drop based off geometric features of the heatsink for varying flow rates. The Matlab code associated with the following figures can be found in the Detailed Design Documents folder. Following the code, we produced the figure below of static pressure drop and average velocity.

Pressure Drop and Average Velocity Across Fins

To understand the effect of the impedance curves, they need to be superimposed onto the fan performance curves. Picking a few points around where it crosses the performance curve and generating a line yields the operation point. The figures below show the operating point for our mock-up fan and heatsink and the operating point for our chosen fan and heatsink. In both cases, curve 3 is the performance curve for the fans we chose. The mock-up heatsinks were 9.4mm tall fins whereas our chosen heatsink will be the same size with 12.7mm tall fins for a greater surface area.

Operating point of ~1.2CFM for mock-up fan and heatsink

Operating point of ~6.2CFM for chosen fan and heatsink

The housing was designed to have a sleek appeal and generally to keep the weight and size at a minimum. It also was designed to keep the components snug and safe, however testing and printing of the housing will determine if these are correct. Below is our view of what the device will look like as well as an exploded drawing of the full assembly with all components.

Preliminary Housing Design

Exploded View Drawing of Full Assembly

After our preliminary review and adding the other necessary features to the design, a battery compartment was installed along with the loops to hang the device and holes for electrical interfaces with the user. Grates were also added to the fan inlet and hot side output to prevent consumers from touching the rotating or hot components. Stress analysis was run on the loops for the hooks in Solidworks to determine the areas of high stress concentration and displacement when force is applied. By angling them toward the user, these are minimized and the highest stress concentration can be seen in later figures near the portion where the chain would be attached. The displacement was found using both Glass-filled Nylon 6/10 and ABS. All figures are with a 45o force of 9lbf or ~40N. This displaces the loops by less than 0.005mm. Below are some images of the final design.

Full Final Assembly

Full Final Assembly Exploded View

Section view showing battery compartment

Stress Concentrations in the Loops

Displacement of ABS Loops

Displacement of Nylon 6/10 Loops

As we increase the flow rate and velocity however, the temperature at the outlet will also increase. This is shown by the ANSYS figure below. It assumes that the thermoelectric temperature and fin temperature remain constant(based on our mock-up data) which may not be the case for increased flow. The left, top contour is with boundary conditions of constant temperature and flow at 1 m/s. The left, bottom contour is with the same boundary conditions but a flow at 3 m/s. The right contour uses a heat flux boundary condition to show the temperature distribution along the fins at a flow of 1 m/s.

Temperature Contours from ANSYS Fluent

To better understand the whole system interaction, we then ran the negative space of the system to help validate our mock-ups and analysis. Using similar methods with the constant temperature boundary conditions, the inlet velocity was determined by breaking down the flow rate for a 40x40mm area (the size of the fans). The solutions were run to 200 iterations to save time and by around that point, they had reached a steady-state value. Below are some of the results.

Temperature Contour of Full Assembly with two different Thermoelectric Temperature Gradients

Velocity Vectors throughout the device

Pressure Drop across device for 6CFM fan and mock-up 1.2 CFM fan

Temperature throughout device with different thermoelectric gradients and fan inputs

Velocity for 6CFM and 1.2CFM fans

These results correlate well with our mock-up testing of a 3D printed initial design. They also give us validation that the thermal portion can produce just under the requirements if the temperature gradient were to remain constant, however we believe it will slightly increase pushing us past our requirement.
Overall, the breakdown for the sub-system with regards to the engineering requirements is shown below.
Engineering Requirements Required Value Calculated Value Validation SME Approval
Weight 4 oz. 5.57 oz. Datasheets and Solidworks model -
Volume 15in3 13in3 22.35in3 with battery Solidworks model -
Cost \$6 \$14.02 BOM -
Power - 3W Calculations -
Velocity of Air 1.5 m/s 7.5 m/s Simulation and Calculations Day

Thermal Subsystem

A different thermoelectric was purchased and a third mock-up revision was run to gain some more understanding. The data collected as also used in the ANSYS Fluent modelling. The thermoelectric ran at 6V and 1.13A. The fan ran at 5V and 0.188A. Below is a table of the rest of the data compared to previous mock-ups. We also got a rough measurement of the output temperature just before the nozzle exit, as the thermocouple was greatly affected by ambient temperature. Our measured air temperature coming out of the nozzle was 60oF. For the fourth mock-up, the thermoelectric was not working for some reason, therefore no temperature data was recorded but airflow was.
Mock-Up Fan Velocity Exit Velocity Ambient Temperature Cold Side Temperature Hot Side Temperature
Initial 2.5 m/s 0.24 m/s 71oF 74oF 121oF
Second 2.5 m/s 0.45 m/s 75oF 64oF 108oF
Third 2.5 m/s 0.45 m/s 71oF 57oF 102oF
Fourth 2.5 m/s 1.2 - 1.4 m/s N/A N/A N/A

For the thermal subsystem, once pressure drop was known, modelling was done of various heatsinks similar to our selected one for the mock-up and for our chosen heatsink. This was done in ANSYS Fluent using Solidworks models as the geometry. It's important to note that Fluent models the 'solid geometry' as the fluid space. Therefore, the actual solid model needs to be converted to the negative space. For meshing, with all the fin boundaries, a body mesh was used with a size of 1mm for the elements. The boundary conditions considered the fin walls and bottom of the fins (top surface of the heatsink base) to be a constant temperature. This assumption seems reasonable as the small fins will gives us a near 100% fin efficiency, and the enclosed space and small fins allows for the thermoelectric temperature to conduct to the edge of the fins. One of the boundary conditions was tried with a heat flux of -200 W/m2 to determine the temperature distribution along the fins. The inlet velocities were mainly constant at 3 m/s with a few run at 1 m/s for the 9.4mm tall fins and the temperature at ambient. The solutions were run to a converging criteria of 1E-4 for continuity and x,y,z momentum equations and 1E-7 for the energy equation. The flow was modeled as laminar because of the low Reynolds numbers in the fins. The solutions didn't converge but reached a steady-state and so were only run to 400 iterations. Below are the pressure and temperature contours for the various models used.

Pressure Contour for 9.4mm fins

Temperature Contour for 9.4mm fins

Pressure Contour for 12.7mm fins

Temperature Contour for 12.7mm fins

Generating a line in between the fins during post-processing allowed graphs to be created to show the pressure drop, temperature, and velocity inside the heatsink. These can be found in the attached Excel document. Below is an example of one of the graphs. For the temperature and velocity, the elements were too big such that the boundary conditions mess with the results. A more refined mesh will be created to better represent the data between the fins.

Pressure Drop from ANSYS data for various heatsinks.

Future analysis is to include the refined mesh for our current mock-up and selected prototype heatsinks. We also plan to create a model for the entire system to observe the flow and temperature within the housing.

Overall, the pressure data from ANSYS, Matlab, and our mock-ups seem to agree for the most part. We are confident in our fan selection and the temperature output. The subsystem mapping to the engineering requirements is shown in the table below.

Engineering Requirements Required Value Calculated Value Validation
Weight 2.5 oz. 2.2 oz. Datasheets
Cost \$9 \$13.54 BOM
Power - 9W Calculations
Temperature Differential 10oF 11oF Mockup

Electrical Subsystem

Original Electrical Subsystem
The Electrical Subsystem charges the battery and transfers energy from the battery to the other subsystems it also provides user input. The subsystem overview can be seen below. This was the original design. The updated design can be seen further below.

Electrical Subsystem Overview

The battery is not included in any engineering requirements. However, we wanted to decrease the power requirements of the system as much as possible to enable our customer to create the smallest custom battery possible. The power calculations for our system can be seen here where battery capacity is determined using lithium-ion batteries.

The boost converters must convert the voltage from the lithium-ion battery at 3.0 to 4.2V to a voltage that can be used by the other subsystems. The image below shows the circuit for the Peltier boost converter, converting to 6V for use with the thermoelectric. For our design, the thermoelectric requires a current over 1A. In order to keep costs down, a boost converter IC with an external MOSFET is used.

Peltier Boost Converter

This circuit was simulated in PSpice. The input, output, and switching voltages can be seen below. The worst case scenario input of 3V was tested.

Peltier Boost Converter Simulated Input, Output and Switching Voltages

There are losses in the system as seen in the following image. However, with smart part selections, the efficiency of the converter can be increased.

Peltier Boost Converter Efficiency

A thermal analysis was also run on this boost converter due to the higher currents to determine if it would affect temperatures of the entire system. As seen below, the effects are minimal.

Peltier Boost Converter Thermal Analysis

The image below shows the circuit for the Fan Motor boost converter, converting to 12V for use with the thermoelectric. In contrast with the first boost converter, a large current is not required to run the fan motor. Therefore, a simpler circuit can be used.

Fan Motor Boost Converter

This circuit was simulated in PSpice. The input and output voltages can be seen below. The worst case scenario input of 4.2V was tested.

Peltier Boost Converter Simulated Input, Output and Switching Voltages

There are losses in the system as seen in the following image. However, with smart part selections, the efficiency of the converter can be increased.

Fan Motor Boost Converter Efficiency

With these losses in mind we can calculate the power and battery required for the system to run for one hour continuously seen in the table below or here. From this information we have decided to go with a 4000 mAh battery. Due to size restrictions we elected to use two 2000 mAh batteries in series. The model of the battery can be seen below.

Electrical Power Calculations

2000 mAh batter with protection circuit and thermistor

Updated Electrical Subsystem
Our design was updated to provide better efficiency and lower the unit material cost. The updated design uses two lithium ion cells in series when the device is in use to supply the same current at double the voltage previously. The updated power converters can battery management circuits can be seen below.

Thermoelectric Buck Converter Circuit

Fan Motor Boost Converter Circuit

Battery Management Circuit

The customer will also be providing a custom battery over the summer to be implemented into the final design.

Drawings, Schematics, Flow Charts, Simulations

• The datasheets for major components can be found here

Bill of Material (BOM)

This project has three separate bill of materials. The Prototyping BOM shows only the quantities and costs to create the prototype. The Project BOM shows the total costs of the entire project, what has been purchased so far, and is used for budget tracking. The Manufacturing BOM shows the cost for the customer for manufacturing and is used to show that the project meets the unit material cost and weight engineering requirements. All three can be viewed here. Below is a summary of the current cost, weight and budget as reflected in the BOM.

BOM Summary

Injection molding calculations can be found here

Test Plans

Here is an overview of the test plans we will preform in MSD II. The procedures and data sheets for all of the test plans can be found here.

Test Plan Overview

Risk Assessment

Our updated risk assessment is as shown below. The file can be found here.

Risk Assessment

Presentation

The presentation can be found as a PDF here or Powerpoint file here

An action item we received from our preliminary detail design is to cut the battery life to ~1 hour and include it in the prototype.