Test equipment (part 2)

This is very much a natural continuation of the previous post, in which the probe points of interest were identified for each of the three planned rakes. The last thing I did was pull out the velocity vectors and total pressure values for each of the points from CFD, and now its time to use them. 

Step one was to pull out the coordinates of the points and put them into CAD. Each point then had a line drawn, extending rearwards by 70mm (the length of the Kiel probes) at the angle of the local velocity vector. One probe was registering a velocity with positive (albeit a very low magnitude) Z component (i.e. moving forwards relative to the car); rather than inverting the probe at this location, it was set pointing straight forwards, with the decision that any reading close to or less than 0 would be considered an accurate correlation for this point. From here, the rearmost point of each probe was joined with a pair of splines to create a 20mm ribbon structure, forming the rake. The choice of connections between each point was a compromise between minimising aerodynamic blockage and structural rigidity. Avoiding excessive blockage consisted of using the minimum number of connections to pick up each probe, while avoiding sharp acute angles and more than 4 connections at each node/probe point. Structural strength came from maximising triangular shapes were possible, or otherwise avoiding parallelism to maximise resistance to skewing.  Figure 1 shows the probe vectors, ribbon structure, and final 3D rake structure at 1.5mm thickness for the straight airflow Z0.35 plane. You can see how the flow is pulled in behind the tyre from the angle of the probes and rake structure at the outboard edges.

The proximity of the two centre ribbons isn't ideal and significantly reduces local porosity, but the alternative was to have unsupported sections and/or a very wavy single centre ribbon that risks adding blockage anyway from uncertain local flow angles and rapid changes in curvature. On a side note, I've been told by three separate people that this particular rake looks like a cat. 

After making rakes for all three cases (two straight and one corner), they were all printed in individual 150mm x 150mm sections before being glued and taped together. The mounting method had been finalised by this point (discussed later in this post), and was such that no additional changes were required to any of the rakes to support it. It was at this point however that I realised I really hadn't planned how to mount the Kiel probes very well, which seems like the most obvious mistake. I should have incorporated mounting holes at the nodes of the structure; instead the probes will have to all be mounted slightly offset from the intended position, and held with (hopefully) temporary glue. This wasn't ideal, but was mostly resolved by adjusting the probe points in CFD accordingly. This meant the positions weren't exactly where we wanted them, but at least we would still get equivalent/comparable data for the positions they ended up in. After initial construction, the rakes were stiffened with a layer of carbon fibre (Figure 2). The carbon significantly boosted the rigidity as hoped, to the point where there was no noticeable flex under 2g of acceleration. It's not the prettiest construction, but that can be solved with a bit of sanding and black paint once we're done testing to make a nice part we can show off to people. 

To figure out how to mount the rakes, the CAD models were imported into the full-vehicle CAD model (Figure 3). It was here where we noticed a missed opportunity to mount the straight airflow rakes on the RHS of the car, which would allow the front rake to share the same mount design as the cornering scenario rake. No great loss as the mounts are easy to make. 

All mounts were designed from 3mm sheet aluminium as a simple material to work with. The front two mounts fasten to the underside of the floor, and with no solid floor behind the firewall, the rear mount fastens to the left upper seat mount. The non-fastened end of each mount sits snugly over the chassis, held by a cable tie and a thin sheet of rubber bonded to the mount to avoid lateral slip on the chassis tubes. The chassis mounts were interfaced with the rakes by three 100mm bolts. By clamping the rake between two nuts on each bolt, the angle could be easily adjusted to match the probe locations to CFD. The final mounting setup provided impressive rigidity in vertical and lateral directions, with longitudinal rigidity provided by applying forwards tension with a cable (to counteract drag forces). The sensor PCBs were mounted rather awkwardly next to the driver, with the requirement of being close to a CAN breakout point. The PCB mounting will be much improved for future tests, with a single smaller board and the addition of 90-degree adaptors to avoid the pneumatic tubes protruding into the cockpit. The Z0.35 straight airflow rake mounting is shown in Figure 4, and the sensor board mounting in Figure 5. 

Before you ask, no, there isn't a gaping hole in the bottom of the cockpit. There's a carbon cover that mounts over the shocks, rockers, and datalogger, keeping to the cockpit closeout regulations. 

We also have a somewhat more professional yaw probe now (i.e. not a piece of wool) which will be put to use at the upcoming track test, but I'll cover that in the next post to keep this one focused on the rakes.