This blog is for entertainment purposes only, and is not meant to teach you how to build anything. The author is not responsible for any accident, injury, or loss that occurs as a result of reading this blog. Read this blog at your own risk.

Wednesday, April 04, 2018

Ch 23 - Oil cooler fan - part #1

Putting "experiment" back into Experimental Aviation

Well, I wasn’t sure where to file this story, but since it had to do with the engine, I figured I’d go with chapter 23, while that’s still available. 😉

So, here’s the story… Last summer I was finally able to convince my wife to fly with me to an event that would have otherwise taken two full days of driving. Flying there and back only took 3.5 hours.

Gina's first Long EZ flight

This turned out to be a fantastic fist flight for her, and really highlighted the utility aspect of the Long EZ. Being able to get off the hot ground, and climb until the temperature was “just right”, really drove a lot of points home.

Panoramic view from the front office window (pre-EFIS cockpit)

However, that is only part of the story. The interesting bit for me happened before taking off to go back home.

First though, let me tell you about the oil cooling system on Terry’s  Long EZ. All pretty standard stuff really, though Terry chose to install the oil cooler on the bottom cowling facing down. 

I have tried blocking 1/3 of the oil cooler trying to keep the oil warmer in the winter.

I am not a huge fan of this setup, mostly because it disrupts the airflow under the cowling. Airflow separation, drag, and turbulent flow into the propeller are a very realistic possibility.

I have to admit though that it is very effective at cooling the oil down, and I have never had any issues with oil temperatures in flight, even in 100ºF days (38ºC). 

Even with 1/2 of the oil cooler blocked, this oil cooler outperforms the engine.

As far as I am concerned, this is a very big deal, so I will forgive it any airflow separations in order to be able to climb on a hot day.

However, in order to get to the climbing bit one has to first reach the end of the runway, and this is where one the Long EZ unusual features comes into play. With the engine at the far end of the airplane, there is virtually no airflow into the NACA vent and into the oil cooler until the airplane starts down the runway.

This has never been an issue for me, even in the hottest of days, but then again I normally takeoff within ten minutes of engine start. What would happen during an extended ground delay? Think Oshkosh departure.

And this is where our short story picks up again… Being at an international airport means that you have to get in line with everyone else, for however long it might take to get everyone in front of you airborne. In our case I was taxiing behind 4 business jets and a King Air, and wouldn’t you know ATC issued a stop to all IFR departures, for an unknowable amount of time that eventually ended up approaching one hour!

Unable to shutdown the engine (a clearance could arrive at any time), and blocked by a long line behind me, this is when my oil temperatures started to climb. Eventually they got to 240º (5º below max temperature), when finally ATC shuffled the big jets around to let the VFR departures go (smart move, albeit slow)

After takeoff all was well again, and there were no further issues. Nevertheless, this occurrence brought to light an inherent weakness, not so much of the design per se, but of its suitability to IFR operations, and their ever present, impossible to predict ground delays.

So I decided to elicit the experimental portion out of Experimental Aviation, and conduct some testing of my own to see if the situation could be remedied. My idea was to add a fan on top of the oil cooler to be powered during ground operations only, then await the hot season and test it out.

I needed a fan that would be able to operate in a hot (perhaps even wet) environment, and it had to be as small and as light as possible. I narrowed it down to this $90 Champion brand, 6.5” (16.5 cm) automotive unit, drawing 6.5 amps at 12 volts.

Here's what Terry lovingly refers to as my "battery draining device". He might just be right.

A few specs in case you were interested

Problem was that the fan is round (duhhh!) while the oil cooler is rectangular. I needed a shroud that went from rectangular to circular, while also translating forward 0.5” (13 mm), and to the right 0.5” in order to have a little more clearance from the engine baffling.

Not knowing how to make it happen in the physical world, shy of removing the oil cooler off the plane and bring it home for some tricky 3D glassing, I started by recreating everything up in CAD in order to at least be able to see a rendering of it on the screen, rather that just with my mind’s eye.

With a bit of CAD trickery I was able to make the shroud appear between the fan and the oil cooler

Removing the fan to reveal the internal shroud air channel

Here's what the shroud should look like

I know what you are thinking... "Good luck making one of those!"

Doing so allowed me to grasp the complex shapes of this shroud, by turning it ever which way on my computer screen.

Then a crazy idea occurred to me, what if I 3D printed the whole shroud, flanges and all. This way I could not only see it on the screen, but also touch it, and play with it, and perhaps some better ideas might start to percolate from all of this.

As luck would have it, the shroud would be just within the maximum dimensions of the bed of my 3D printer, and so a few hours later (30 😬) I was holding a complete shroud in my hands and… she looked beautiful

Small separation on the left side (pre 3D printer enclosure days)

Measure 1000 times, print once!

"Man, the fit on this thing is just perfect!"

As seen from the oil cooler

Forgive me for gloating, but this is some cool shit!

As I expected, the physical model highlighted a few flaws in the design that would have to be fixed in the final version, like the inability to insert the bolts in the bottom flange hole due to the shroud’s wall angle, and the need to have relief for wrenches to get to such bolts. However that didn’t need to stop further testing of this proof of concept, so I took it to the hangar for a second round of fitting tests.

Like a glove! 😀

This gap with the oil cooler (front and back) will be addressed in version #2

The fit turned out to be almost perfect, so I zip-tied the shroud/fan combo to the oil cooler in six places and ran the engine, and the fan without the top cowl. After seeing that the shroud was strong enough for ground testing, I installed the top cowling and ran the engine and fan some more.

Unfortunately the oil temperature never reached above 150º, so the oil was mostly bypassing the oil cooler during the test. 

Obviously this test was inconclusive, and I will have to wait until the summer in order to get the oil to heat up to over 200º on the ground, then run the fan and see if it helps with cooling or not.

While the jury is still out on that verdict, that doesn’t mean we couldn't continue to develop this experiment in order to be ready when mother nature is. So, I started to think about the logistics of how to reproduce the same shroud in fiberglass, and that’s when I had another crazy idea… What if I 3D printed the hole inside the shroud instead, and laid fiberglass over it?


However one last logistical problem was threatening to scrap it altogether. The shape of the shroud was such that removing the mold from either side would be impossible.

After one last head scratching session, I decided to adapt another concept I had already used for my Tormach TTS tool holders.

Interlocking shapes scheme I used for my mill's toolholders

The plan was to break up the mold of the shroud’s hole into five interlocking pieces, and print them together as one unit. After glassing, the center piece would be removed first, then the rest of the mold could collapse inward, and be trashed or even reused.

3D printing all five parts of the interlocking mold together

Hot off the press!

The "Transformer" mold!

I would glass the two flat flanges separately from the main body of the shroud, then join them.

Masking the mold before glassing

Making sure I'd get some release later

Getting started on the flanges

4 plies of BID

About the same here

I used West System epoxy here because I had some left, but I ran out after this piece.

Both flanges cut and bolted to their 3D printed counterparts

Testing the round flange on the actual fan

Finally some glassing action

This Easy-poxy has to cure overnight before one can work with it

Here's the reveal the next morning

Everything going as planned

Wow, this was easier than I had thought!

All the elements of the shroud

This is how they should go together

I will show you how I did that in part #2.