Source: 2021 Q2 Beartracks, Jared Yates
Engine cooling is on the short list of things that builders have ongoing problems with. It doesn’t have to be this way, but I’m not really one to talk. Last week on a hot summer afternoon, we flew our Bearhawk 20 miles north for a photo mission and ice cream run. We shut down long enough to eat ice cream and visit for a few minutes, then started up to fly home. This is the worst case scenario for cooling, since the cowl temperatures rise after shutdown. While enroute, the engine is making lots of heat, but there is airflow to remove that heat. Once the plane is not in the realm of flying speed, the residual heat in the engine parts gets transferred to the air inside the cowl, and everything except the engine gets warmer. After departure and reaching 2000’ AGL, our highest CHTs were up to 425 and climbing.
The challenge with this scenario is that reducing engine output is one way to reduce the temps, but that also reduces the airspeed, which reduces the cooling airflow. I know from experience and education that cooling improvements are incremental. One silver bullet seldom solves the problem, but if I’m going to fly to Oshkosh this year, I need to start incrementing.
Cooling effectiveness feels like black magic, but as I wrote about in the 2015 Q3 issue of Beartracks, it is fairly easy to measure and quantify with a home-made manometer, followed by noting CHTs in flight after making a change. If one wants to burn the fuel and go to the trouble, it is possible to set up the manometer, fly to get a baseline, make a mitigating modification that is supposed to help lower temps, and then fly to see if the pressure differential changed.
As I prepare to start this process, I’ve been collecting data about what those mitigations are. First, we have to ensure a good seal between the baffles and the top cowl, or do like Rob Caldwell did and install a plenum:

One way to study this seal is to watch the wear pattern where the rubber strips rub the aluminum. Another is to use a flashlight to check for leak areas. I feel like in my case I have this worked out pretty well, so on to the next step.
Perhaps the two biggest variables are the inlet and outlet areas. The inlet areas are fairly fixed unless we start designing our own nose bowl openings, but we see lots of variation in the outlet size. This photo shows Ray Strickland’s O-540 powered Bearhawk:

He has a large area, along with a pronounced lip to help reduce pressure. In my case, I’m not running such a large opening, though I feel like the maximum size of the opening has been limited by the way I’ve tied in the fiberglass airbox fairing. If I move the trailing edge of the cowl forward, it will diminish the amount of aluminum left to hold the lower cowl together, but maybe this is what I’ll need to do.
That brings us to the next mitigator, which is the lip. The lip is most commonly composite or aluminum. Starting with a soft 5052 or similar aluminum, the lip can be made with a shrinker and stretcher. That type of lip will also add strength to the lower cowl, just as flanging a rib adds stiffness. It has been proposed that the upward flex of an un-lipped cowl in this area might actually lead to an automatically-reduced opening size, which would be most unfortunate. What is the optimum length of the lip and angle? I can’t tell you, but it sure would be interesting to have some data. For now the best I can do is study the sample of available planes, and hope that the Type-Certificated crowd is flying behind an R&D department that considered all of the possibilities. Intuition and observation seem to point to around 2 inches of material at an angle of around 30 degrees.
All of the above mysteries about cowl lips are moot with the next option, which are adjustable cowl flaps. In some ways, we could think of the lip itself as a cowl flap, but that it’s always open. When enabling the option to adjust the opening, the design considerations change. Builders have executed a variety of different designs. Jonathan Battson made two openings, roughly 5×8 inches, but says he’d rather have them 30% bigger:

He used a piano hinge at the leading edge and made the flaps out of .025 aluminum. When open, they are deflected around 4 inches at the trailing edge. Jonathan uses an adjustment knob to manually operate the flaps and finds that he can reduce CHTs by 30 degrees, though he finds that he usually sets them either fully open or fully closed. One of his primary concerns was being able to keep the engine warmer during long descents. And since engine cooling makes drag, it makes sense to only produce the cooling and drag when the engine needs it. This is a big advantage of adjustable cowl flaps. Here’s a video of Battson’s cowl flaps.
Jon Wheeler used a similar concept, but used input from Jonathan to increase the size to 8.25 inches long, 6.5 inches wide, and 4.5 inches of deflection. Jon used electric linear actuators (Actuonix P16-100-64-12-S) instead of a manual cable. Jon also moved the flaps further up on the cowl, closer to the area where builders have also used louvers.


Speaking of louvers, those are a common option, especially on airplanes with the bigger engines. Lots of folks use salvage Piper parts (PN# 87405-802 (left), PN#87405-803 (right)) with the intent being to further reduce the air pressure below the engine. Jon Wheeler’s implementation of cowl flaps certainly approaches the function of the louvers and how they exploit the relative low pressure area there. Builder Bobby Stokes was having trouble with cooling, so he did some tuft testing and manometer testing. In his louver installation, the cooling and differential pressure were both worse. His louvers are much higher up on the cowl than many others, and there has been some discussion on the forum about which direction the louvers should face. Bobby did find that the fixed cowl flap or deflector at the bottom made the largest difference. He made a version that was adjustable, but didn’t find that the adjustment was necessary and has since gone back to a fixed flap.

Another common mitigation is to round the angle that the exiting airflow experiences at the corner between the firewall and the tunnel. I added a fairing here a few years ago, made from .020 2024-T3, bent on a slip roller. The idea is that this radius fairing helps reduce airflow separation right at the point that shows up in silver on the picture of Ray’s airplane above.
Next quarter I’ll have more news, either a tale of what didn’t work, or more hopefully, what did.