fe1rx adds Brake Cooling Ducts

[fe1rx]

Wanting to add larger dual oil coolers, I realized that these would conflict with the OE brake cooling ducts. Brake cooling being a definite requirement, that got me looking at ways to relocate the brake ducts and preferably improve on the OE brake cooling.

It is easy to look at the OE brake ducts and to conclude that they are probably not very effective. They are deeply curved and don’t look like they would direct cooling air very effectively. Not wanting to simply operate on that assumption, I used a shop vac as a blower and connected it to the air inlet. Wool tufting at the outlet sticks to the boundary of the stream tube (air jet) exiting the duct, clearly revealing its extent.





I was surprised to find that the OE duct actually does produce a coherent stream tube generally directed at the center of the hub. I also concluded that the deep bend in the OE duct functions as an inertial separator, knocking entrained dirt, sand and water from the air entering the duct. This is a good idea for a street car but will come at some performance expense.

By eliminating the inertial separation feature and by moving the inlet down and inboard, I was able to create a 3” round duct geometry that eliminated the conflict with a larger oil cooler and provided a stream tube aimed directly at the center of the hub at ride height. The duct has two small bends with a total combined bend angle of only 45 degrees. I made the initial mockup using some 3 inch mandrel bent stainless steel I had on hand.



This geometry provided an air stream at least as effective as the OE duct.



Some modification to the radiator support was needed, and a couple of jigs were used to ensure that LH and RH sides were true mirror images. The ducts themselves were fabricated from 3” round mandrel-bent aluminum tubing, TIG welded with one 45 degree bend being sufficient for each side. The exit geometry was configured to prevent any tire contact on steering.





The wheel well liner required modification due to the relocation of the duct exit (green being original location, image is LH wheel well liner looking forward).



The inlet bellmouth required significant effort. A radiused lip on all air inlets is desirable to minimize flow separation at the inlet. Any flow separation will reduce the effective inlet diameter and reduce the flow. Aligning the inlet with the air flow direction is also beneficial. Unfortunately the air flow direction is unknown and is itself affected by the presence or absence of the inlet. Some CFD might be useful here. We know that off-center the air flow will have some outward component. Tufting on the front of the vehicle shows this clearly.



Accordingly I rolled the dice and chose to orient the duct such that it points inboard 15 degrees from straight ahead, and to create a bellmouth geometry with a more generous radius on the inboard side. This kind of approach can be seen on the BMW M3 GT4 and other true race car installations (unlike the highly stylized and exaggerated funnel designs found on road cars).



The bellmouth was fabricated from annealed aluminum, with the three components hammer formed over forming blocks, trimmed then TIG welded.

In order to see if my assumptions worked out as intended, I installed a GoPro in my wheel well, tufted the outlet and went driving. I was unsure how much the general turbulence in the wheel well, and the rotation of the wheel would upset the stream tube. The following image is a screen capture of the flow at 100 km/h.



The flow is strongly directed exactly where I intended, with some attraction of the flow to the rim of the wheel due to the Coanda effect. My brake scoop appears well located to assist directing the flow to the center of the hub, so I am pleased with the result. I intend to do some tufting on the front of the vehicle to see how the air is approaching and entering the inlet, but that is largely an academic exercise.

If it were not so difficult to accomplish with our vehicles, I probably would have considered a fully ducted installation using flexible ducting. 3 inch fully ducted installations do appear to be the de facto standard. Such installations invariably have many bends which will result in some duct losses. Also, the accordion pleating of typical ducting will result in significant turbulence in the flow. Terry Fair from Vorshlag is a proponent of deflectors, or (as in my case) ducting feeding deflectors as opposed to fully ducted installations. Based on my experience, I am inclined to agree with him. My design provides a very straight path for the air and should have minimal ducting losses, provided the inlet flow is well aligned.

[fe1rx]

I found some comments from Terry Fair on the Professional Awesome FB group after I was fully committed to the semi-ducted approach, and this link gives a description of the evolution of his thought on the subject which you might find interesting:

https://trackmustangsonline.com/thre...-2#post-250600

[fe1rx]

I have tufted the inlet and front fascia to see how well attached the flow is in the duct and how much disruption is caused by the ingestion of the lateral flow across the fascia. I used a GoPro mounted ahead of the vehicle to get a good view of the inlet under driving conditions. The image below is at 100 km/h road speed. The lateral flow is clearly visible on the fascia. The flow inside the duct is very well attached so the bellmouth seems to be working exactly as intended. I believe the large radius on the inboard side of the bellmouth is essential to accomplish this.