The break in my Ironman training due to this relentless virus has provided me with some needed time to tackle the TCR1 frame and fairing design. Ben Eadie is taking my basic geometry concepts and re-modelling them in an application called Solidworks which will produce accurate drawings for me to use to fabricate. It's much better than the 3D modelling software I use and it does some basic finite element analysis to calculate the optimum frame geometry and tubing sizes. In the end, this is going to make my fabrication job MUCH easier because Solidworks will generate proper dimensioned drawings.
Before I build the new TCR1 frame, I thought that I should finish a fairing design first to make sure that the frame works properly with the fairing. It's a good thing I'm doing this because I once I started experimenting with some fairing shapes, I realized that I needed to raise the seat quite a bit to see over the fairing. So, I increased the seat angle, and then raised the pivot.
The Old Geometry:
The New Geometry:
I spent a couple of days working on a laminar flow fairing design which resulted in two possible choices for a fairing. Here is a quick tutorial on how I designed the fairings:
1. To calculate the drag for a basic fairing shape, I wrote a little software application called "ShapeViewer" that allows me to import a photograph, then trace around the image and export the resulting outline to a CFD (Computational Flow Dynamics) program called "DesignFlow" by Dreese Software. DesignFlow simulates airflow around the outline and generates a pressure diagram and resulting drag values for the airfoil shape. ShapeViewer allows me to trace around the basic outline of various shapes like a bird, shark, or existing HPV fairing to determine how aerodynamic those shapes are. Then I can modify the shape and re-test with DesignFlow to optimize the aerodynamic efficiency of the shape. If you are interested in obtaining a copy of ShapeViewer, send me an email.
ShapeViewer Screen (click for larger)
DesignFoil Screen (click for larger)
2. I tested about 20 different airfoil shapes including most of the HPV Streamliners I could find (Varna Diablo, Mango, Baraccuda, etc), various fish, birds and some existing airfoil shapes. With DesignFoil, I calculated the drag for each shape for three different conditions: First, a simple, clean airflow, second, airflow at a 5 degree angle which would simulate the effects of a cross wind, and third, a simulation with insects effecting the leading edge. The table showing the drag values for all 20 shapes is here:
Airfoil Drag Values
Then I sorted the results and looked for the lowest average drag for all three conditions. Then, out of the top three (the three lowest drag values), I selected a shape which would be the easiest to build a fairing out of - ie: room for my legs, feet, shoulders, etc. The airfoil I selected is an aircraft airfoil called a Griffith wing. It's about 75% laminar, but still has low drag values in a cross wind and with dust or bugs disrupting the laminar flow on the leading edge. On the following DesignFlow wind tunnel simulation plots, the red line is the airfoil shape, the green line is the airflow, the red dots show the extent of the laminar boundary layer, and the blue line shows the pressure coificient.
Griffith wind tunnel plot
The second airfoil shape I selected is one based on Jeff Wills "Lunatic Fringe" streamliner. I modified it somewhat to optimize the amount of laminar flow (100% if that's even possible!) and to somewhat accommodate the enclosure of me and the frame. I call it "LunaticMod".
LunaticMod wind tunnel plot
3. To construct a model of the fairing based on the airfoil shape, I imported the basic shape into Strata 3D, then stacked 5 ribs over the frame that would contain both the frame and my body. It is important to note, that the airfoil shape cannot be scaled unproportionally - it MUST be scaled evenly width and length or the airfoil dynamics will completely change. Also it is important to note that as the scale gets smaller (or larger), the renolds number will change. I won't get into what a renolds number is, but basically, the original shape is optimized and sized for an exact dimension (ie: length of the fairing) and approximate speed. That said, I ran some tests for a variety of renolds numbers on the airfoil shape and the resulting drag numbers and pressure flow plots were not different enough to justify an entirely new shape for every scaled version - so, I can scale the airfoil shape up and down, as long as I scale it proportionally.
Note the airfoil shapes stacked, scaled and aligned to fit over the body and frame.
Then the ribs are 'skinned' or joined together to create 1/2 of the fairing. The fairing half is then mirrored to create the entire fairing.
The "Griffith Fairing" is a more robust, partially laminar design and the "LunaticMod Fairing" is extensively laminar, but more sensitive to cross winds, bugs, and road vibration. I'm not sure which fairing I'll actually build - possibly both.
Here are some 3D renderings of the fairings. Click on the thumbnail to see a larger image.
To receive these daily reports by email, click here.
Click here to go to the HOME PAGE