Cooling

Date:  03-14-2025

Number of Hours:  20

Manual Reference:  no ref

With the upper and lower cowl flaps now installed, the large avionics job remained.

After some planning I laid out conduit for wiring so that there would only be one plug for the top and one for the bottom cowl. The the carefully applied paint was taken back to the carbon and I hot glued the channel in place.

After it was fixed I went ahead with micro to transition the edges and properly glue the ‘pipe’ in place. Then a layer of carbon fibre over the top.

The top cowl just needed the one bridging piece for the wiring.

It then got the royal treatment from my wiring expert, definitely not me!

Somehow he put a huge amount of circuitry into the little box. there are eight relays in there! Not only driving the up/down signals but also hooking up to the four flap microswitches and one on the throttle to give auto retract if I forget them at take-off. There is also taking the signal from the flap position indicator and hooking that into the Garmin engine box. A huge amount of work. I just had to secure the box in the hellhole with clickbonds and do a few other ‘homework’ tasks.

Here they are in action. The wiring worked first try which was impressive.

They do look good.

WEIGHT
A total gain of 2.32kg (5.1lbs) incl. all reinforcement and hardware
Before the mods: Lower Cowl 4.14kg (9lbs 2oz)   Upper 3.37kg (7lbs 7ozs)
After: Lower Cowl 5.38kg (11lbs 8oz)   Upper 4.45kg (9lbs 8ozs)

The upper and lower cowl will need a repaint. Partly from blistering when I lost coolant during testing and these big cutouts for the flaps will need some transitioning work and edge rectification to make them pretty.

With a number of other small jobs done, I am ready for slow and fast taxi testing on the runway.

 

 

Date:  03-07-2025

Number of Hours:  110

Manual Reference:  no ref

Time for the upper flaps. I’ve been working 70 hours a week on this for some time and I didn’t expect things were going to get easier. Its just a matter of showing up and keep pushing. If you are thinking of cooling flaps for your Long-EZ, you have been warned.

This is where the top right flap mechanism has to go. Weaved in-between that stuff on the left side. There just wasn’t room on the right side of the opening.

As usual I start with ‘Masonite’ shapes and scrap metal.

That smaller shinny black pipe is in the way of everything. I managed to move it .5″ with different fastener positions. I could just remove it, but then the engine wouldn’t run.

I kept playing with shapes and putting the cowl on and off the plane dozens of times, it feels like hundreds.

One of the challenges is getting the geometry right so the flap opens a certain amount. Then its lots of testing.

All seemed ok so I could go ahead with the  hard point of .25″ marine ply.

The hardpoints get a ply of carbon to lock then in place although the fasteners really do that job.

Given it all seemed to work I made the main pieces out of my 2024-T3 aluminium and tested again.

The opening I want is 190-200mm which is what is happening with the lower flaps. This is not critical and +/-5% will probably not be noticeable.

Everything seems to work.

Then for some reason, gremlins or evil fairies,  it doesn’t quite fit anymore and I’m ‘just’ touching in a couple of places.

So despite making a ‘finished piece’ it was wrong and I got to make another one, after some more prototyping, which is the finished piece on the right.

I got it all working, happy days. So on to the left side upper flap. It has to be easier? Hahaha …of course not.

The obvious spot for this left side was to put the mechanism on the inside to match what I’d done on the right side upper flap. There seemed to be more room too. This was going to be easy.

I made my usual bunch of prototypes and endless off and on to ensure it was going to fit. I probably spent 20 plus hours and got the push arms to fit, but I just could not make the actuator fit in the end. It was just too near the engine mount. This was a fail.

OK, I’m going to have to make it fit on the left side for this left flap. As you can see it just needs to fit around the engine and a few pipes here and there. On the plus side, there is heaps of room for the actuator.

It was getting weird but one prototype gets made, then acts as a template for the next which has ‘modifications’.

More on and off testing.

Getting closer.

This is nearly it.

I did make a nice metal one before this as I had it nailed. Wrong, nearly nailed doesn’t count so here’s the next one after two more Masonite versions.

This is quite pretty, like some strange weapon from movie.

This works. There is a bit of a catch in the door opening and closing to do with the cowl curvature that I will address.

The four flaps now done with a working mechanism. Wooohoo!

These are the prototype pieces that I made.

I don’t use CAD design and CnC cutters or plasma. I use MAD. Masonite Aided Design with a cordless jigsaw, drill, hand files and sandpaper.

I think this is more efficient as I can knock out a part pretty quickly. A final piece takes me around 2 hours and a few minutes for a prototype. It would be seconds with a plasma cutter…and days getting it set up just right and driving to someone’s place to do the job.  Then waiting for it to be done, then finding out its nearly correct and needing another one made. My own gear would be huge time and expense. I need it now!

Next up is to design some sort of flap alignment system as these guys are likely to move around when closing in high winds and they might jam open or at least be glitchy. I’m not having that!

They also need the electronics, which is being worked on.

 

 

 

 

 

 

 

 

Date:  03-02-2025

Number of Hours:  137

Manual Reference:  no ref

Here we go with building a mechanism for opening and closing the flaps. I started with the lower ones.

Criteria:

1. Getting the actuator as far as possible from the hot exhausts.
2. The second big one is trying to fit a mechanism that works around all the pipes and guff I already have under the cowls.
3. Finally I need something that pulls in the best possible direction, upward, for keeping the flaps tight and closed in flight.

Before we get going. Here’s the actuator I’m using. Its rated at 16lbs of force. There are stronger at 38lbs but I think if I needed that it might break the cowl!

Firgelli FA-MU-16-12-2 Linear Actuator

Here’s the first idea. I need the actuator down near the hinge (away from the exhaust) and a scissor action like my original prototype with ice cream sticks.

This angle will help pull the flap closed.

I soon found I needed more angles to pull on a flap arm. I did make a bunch of fittings which took an afternoon and that’s already one of many masonite arms. The fittings in the end were never used in final versions.

Now I’m getting a bit closer working on going around pipes with a more refined piece. Oh and did I mention I was NOT happy with the way the two doors fitted on the cowls. Yes so I cut off the flanges and redid that whole job. You can throw in more than 20 hours extra work that I made for myself. How to do the flanges is in earlier posts so I’ll spare you more pictures of the same process.

I am not seeking perfection, the enemy of good enough, I do need things to fit functionally. There is no compromise in that. The little arms with the rod ends do look nice though?

I did add those two largish switch to the left side console, just above the seat warmer switches. These will be for upper and lower flaps as pairs.

I continued with my prototyping work getting the hinge mechanisms to work around all the obstacles. This is just a fraction of what I ended up with as the prototypes progressed.

Here things are starting to work. I’m up to putting in hard points, 1/4″ marine ply, Fasteners are going through the cowls and are all going to be rock solid. I do not want any of this stuff coming loose and going through my prop. The actuators have 16lbs of force which is a lot of ‘pull’ in one direction.

Many, many days got us to this point. One of the hardest parts is the geometry of the set up. Change one thing and something else is wrong. I spent several days just playing with this on the one flap. Getting it to fit under the cowls needs to be correct within 2-3 mm or it is going to hit something.

Getting to here, just for this one flap was 14 solid days work.

Here you can see it works pretty well off the plane.

That fancy looking arm with the lightening holes is not CNC or plasma cutter made, its old school. A jig saw, hand files, a drill and sandpaper. It takes me around two hours to make one from scratch. It would take that long to do a CAD drawing and then days to get it made. It is so much easier to just knock one out. If it’s wrong, off I go and make another one. For a one off part, I think this is still the best method.

This is on the plane and missing the pipes and so on.

You can see on the lower middle an extra fitting? This is a bracket for a micro switch which will signal if the flap is open or closed. It took a bit of head scratching as usual and it is adjustable of course.

Time for the second lower flap.

You’d think once I’d made one the rest would be easy? Wrong, each location has a fresh challenge and all the geometry is different. This photo is several days in of trying to nut it out. I experimented with multiple locations for the middle hinge point, it was a beast yet to be tamed.

I tried a lot of different ideas to make this thing fit and open up to the same 200mm as the other flap. Within +/- 5% is the goal. They will look weird otherwise.

I’m not there yet but this gives an idea of how much these guys will open.

Part of the process is trial and error to get all the balls in the air at once. Opening size, missing the pipes, fitting under the cowl. Yes this means cowls on and off around 50 times for this one.

Eventually, as in many days later, this has come out looking a bit like the other side. In the small details it is totally another animal.

This side also needed the flap micro switch for open/closed indications which is almost finished in the picture.

I’ve added a RayAllen flap indicator. This has a 1.2″ travel so I have ‘geared’ it for the 2″ actuator travel. We will hook it up to the G3X flap annunciator and I’ll be able to partly open the lower flaps in flight and see a position in the cockpit. While the concept and design of these cowl flaps is for ground cooling, I have made them really strong just in case a little extra cooling in a slow climb would be handy on a very hot day.

I thought this lower cowl flap process was going to be the most time consuming and the most difficult compared to the top cowls. Of course I was wrong and another nightmare awaited me.

 

 

 

 

 

 

Date:  02-09-2025

Number of Hours:  52

Manual Reference:  no ref

With the flap doors and lips made all I need to do is fasten the hinges. Easy right?   ……..not so easy after all.

I used this hinge on the oil door and it worked out really well so I went ahead and bought four of them.

In my confidence, I even put in the hardpoint of plywood and covered it in carbon on both lower cowl openings. I used the same place as the oil door.

There is a problem. Maybe its because the lip is thicker. I spent a couple of days trying everything I could think of. No luck. What did work, doesn’t on these flaps. I was starting to think I needed to make giant ‘Jay hinges’ as people use on hatches.

I made a few prototype of these ‘Jay’ hinges… but they had exactly the same lip problem. It is really the compound curve right at the opening that is the issue.

An engineer friend visited the hangar, looked at the problem and asked for a piece of paper and an old school drawing compass. He made a few squiggles and told me the hinge rotation point needs to be further back, in fact 20mmm from the edge.

I played around with hot glue and the top cowl and that seems like it might work.

So I have ground off my nice hard points and built another one further back. I love doing things twice. I’m just doing one side at a time now. The other big thing is that I’ve cut the lip all the way back to fit the hinge rotation, kind of an insert that you see above.

I’ve tested this and it ‘nearly’ works, like the earlier Jay hinge ‘nearly’ works. I’ve had to trim the edge of the flap right down so its ‘edge’ and I might need to make a slightly larger than ideal gap around the rotation edge of the flaps, but it will work with this amended and rather unsightly gap at the pivot point.

This is a bit of a relief, although not ideal, it is a solution I can live with.

I need the hard point both for getting the hinge line equal on the door and the cowl as well as for strength. I also need to get the hinge pin going in level so that it doesn’t bend which would cause it to wear very quickly. A little micro does the trick here.

Here you can see the gap at the hinging point and how the door ‘just’ misses the edge. I may have to open it up a touch more. There is a bit of sideways wiggle in the hinge that makes quite a difference to the opening gap required. I’m going to get the mechanisms built and see if that stabilises the opening. If not I have a couple of ideas to try.

We do have a door that opens and closes. OK for now. I did start on the mechanism next and that produced a few new problems of course. I’ll finish off the hinge story here and we can have a look at the mechanism issues in my next post.

Meanwhile I am just not happy with how these doors fit against the lip. Everything was pretty good until I put the cowls back on the plane. Once in place, they flexed into a slightly new position. On the plus side the top cowl door lip are usable. The bottom ones need to be done again and set up on the plane when I put the wet carbon fibre down.

I’ve just the lip off although I did leave the thickness as its strengthening that part of the cowl at the opening and I can live with this.

This time I’m setting up the door as flush as I can make it with all the fasteners done up. Of course I had to remove the exhaust pipes so I have access from the top.

All done and peel plied. These layups take so long that I’m getting the carbon wetted out on plastic and then putting it in the fridge. I then take out enough for two piles of the seven and the rest go back into the cold. It is the height of summer here and pretty hot in the hangar. Not as much fun as it might be.

I put the top cowl back on to let it cure. The result is not perfect, but given I’ll have to refinish the cowls again anyway its within ‘range’ of having an acceptable result.

Of course I had to do the whole process on the other side as well up to this cleaned up state.

Again the fit is better than before and all the ‘error’ cost me was several more long days. Every delay in getting to first flight hurts. The clock keeps ticking. On the plus side I am very much in ‘builder mode’ and going in to work every day until it is done.

So the hinge works. Now I have all the hardpoints to do of the other three and mounting of the doors as I fit the hinges. There needs a bit of trimming work to get the gaps right.

Hardpoints on the top as well. The earlier cowl stiffener probably saved me from having to do these twice. It was still a LOT of work fitting the doors with the lips and it is far from perfect as the cowls flexed a little after being cut. It will work out OK with time and care during the refinish.

The hardpoints got a ply of carbon to seal them in. They will get some levelling down the track too. I do need to keep moving however. I want to see how the mechanisms work before I do a final fit as a few more little things might be needed.

I decided to get the four hinges working before the next step so I at least had a feeling of completion of something.

If you stand back far enough these look ok. Just don’t get too close. OK, lets move on and make these doors open and close with electric actuators while somehow fitting under the cowls between all the engine hoses and bits. Wish me luck.

 

 

 

 

 

Date:  02-02-2025

Number of Hours:  62

Manual Reference:  no ref

With the testing complete and the efficacy of the ground cooling cowl flaps proven by the data, one thing remains. I have to build these things into my ‘real cowls’. The last major step before first flight.

Here we go. The Top cowl weights 3.37kg (7lbs 7ozs) and the bottom 4.14kg (9lbs 2ozs) These flaps and the mechanism are going to weigh a ton. I need them strong. It will be added to my Weight and Balance sheet.

Using the dummy cowl cut-outs as a template, I marked out the flap outline on both upper and lower.

I had to be a little more careful than the dummy set. On the top ones I tried to make them pretty even. The bottom set are offset slightly due to the exhaust pipe layout, but only a little.

Cutting them out was easy with a Dremel. Of course these pieces are only the thickness of the cowl and you can cause flexing by just waving them around. They will need some force to ensure they are tight and I do not want flaps broken off and in my prop, that would not be good. This means they need to be rock solid.

If I build them well enough to open in flight they might be useful in a prolonged climb at low speed if there is a cooling issue in that configuration. I don’t need them for speed brakes, if I did the approach is just wrong and I’d need to go around.

All the paint had to come off the underside for the next step.

I need to make ‘hardpoints’ where I can attach the actuator arms to push and pull on the flaps for movement. It must take considerable force. I’ve modelled this on the landing brake construction to some degree as we know that works in flight and OK to use at 90 knots or so.

I’m using the marine ply that was left over from the firewall. It was quite thick so I’ve routed pieces down to .25″ The shape is to allow for the hinge attachment as well as the actuator bracket.

Given the cowl doors are a compound curve, I’ve just placed them on the carbon and filled the gaps with dry micro.

Of course I have to prep the cowls themselves for flap ‘lips’ by removing paint. It took some time as I didn’t want to remove any carbon either.

Yes the top cowls had to be done too.

I made simple carboard surrounds and did my thing with pour foam. This is going to give the flaps a lot of strength without too much additional weight.

Carefully trimmed there is not a lot of foam left. The edges are ready for the carbon to carbon join.

I thought it a good idea to mark out where the hardpoints are. Once it is covered in carbon, I’ll only be guessing otherwise which could be an issue for placing the hardware.

This is 3 plies of carbon later.

It does need a bit of a trim.

These are really solid now, although heavy. Better strong than departing the airframe unexpectedly.

I had to do the same full treatment for the upper cowl flaps from cut outs to this stage, timber, pour foam and carbon. It is quite a lengthy process. Here’s how I do the carbon. Marked and laid up between thick plastic.

Probably vacuum bagging would be one more step for improvement, although the plastic technique gives a really good ‘glass to resin’ ratio. It’s just not something I’ve pursued on this build. I have a ‘fear’ that I’ll ruin the part if I don’t get it just right and this hand layup thing is working out.

Cleaned up and ready for the next step, making the cowl lips for them to ‘sit in’.

Each flap had to be covered in grey duct tape which I still use as a mould release.

The flap is placed as best I could back where it was cut-out from. Hot glue and sticks are my weapon of choice for this sort of thing.

I couldn’t get absolutely perfect alignment. The cowl has flexed somewhat with such a huge holes in it.  Its not too bad and I will have to refinish these cowls at some point after a bit of testing. They were painted so beautifully too.

I will have the opportunity to do a little filling to make them right. It will be another big job. Painting only works if the preparation is ‘perfect’ and that gives you only a good job.

Now for the lips. I have gone for 7 plies. While only 1.5″ wide, it is still a lot of carbon and preparation.

Both sides were done on the same day with peel ply at the end. I nearly ran out of time as the epoxy was starting to go off.

It was a bit of a job cleaning off all the hot glue. The tape worked really well and the doors came out easily enough.

After a trim, I have nice fitting flap doors. Not perfect, but they are close and will, after the repaint, be very nice.

Here’s how it looks on the reverse side. What colour should they be when they stick up in the air?

Here’s the other lower flap.

Of course there was the same lengthy process with the top cowl flaps.

Here’s the completed lip on one of the top flaps. There will be enough pictures later as I work on the actuators.

With the light just right the cowls still look good. Just don’t get too close until the repaint, OK?

 

 

 

 

 

Date:  01-18-2025

Number of Hours:  0

Manual Reference:  no ref

This is the data collected from 11 engine runs in different cooling configurations. Cowls on and off, 4 cooling fans behind the radiators, installed and removed with two bottom cowl flap openings and finally with all four openings and the flaps temporarily installed fully opened.

NOTE: My engine is liquid cooled. My interest is in the coolant temperature, similar to the numbers that you might have in your car. Oil temperatures are also important.  CHT’s have no relevance of course in a liquid cooled set up.

Results
With the four large flaps open and no cooling fans I can stay in the hold for 40 minutes with engine temperatures low enough to safety take off.  (Coolant 192F, Oil 205F)

I can now start the flap modification on my permanent cowls now these ‘test dummy ones’ have given me a proof of concept.

Summary Data Chart

You can see with the cowl flaps open and no fans I am below 200F after 40 minutes at idle RPM and the oil is 205F as well. Safe engine temperatures for a take off. I hope I’ll never have to wait 40 minutes in the hold but now I know the plane can handle it.

You can see with no flaps I would be good for less than 10 minutes, a bit like my old Long-EZ with the air cooled engine. This graph also shows that the addition of fans was a very minor improvement and is certainly not worth the trouble.

The Raw Data
I have a lot more data than this. My G3X record numbers every second! I took notes every 2 minutes in the cockpit. That is plenty. What follows are copies from my excel sheet and may be easier to read in this format. Feel free to get in touch if anything is unclear. Temperatures are in F° time is in minutes.

Time	Test Duration	Coolant F Temp Post	Pre Rad Temp1	Pre Rad Temp2	OIL Temp	Oil PSI	Aux Pump ON	Coolant PSI	Engine RPM	Volts	Amps	COMMENTS
Taxi Cooling tests			Auto Temp set to 170F, Non fan radiator areas gaffer taped on rear side 3 X 5.2″ Spal Puller Fans
28/11/2024   BOTH COWLS OFF
12.51	0					48						Startup
12.56	5	162	169
12.59	8	171	177
13.02	10	174	179
13.04	12	176	179
13.09	17	177	180
13.16	24	181	185				ON	7		12.7	38
13.18	26	182	188
13.19	27	184	189
13.22	30	185	187
13.23	31	184	186					8
13.24	32	183	185
13.26	34	183
13.31	39											Shut Down
28/11/2024   BOTTOM COWL ON TOP COWL OFF
15.22	0	105	104	106				7
15.27	5	175	176	177	159		ON
15.31	9	182	185	185	177	50
15.35	13	190	193	192	196	49		9
15.38	16	195	197	198	206	47		10
15.41	19	199	201	201	211	47		10
15.44	22	203	205	205	216	47		11
15.47	24	208	210	210	219	47		12
15.50	27	209	211	210	222	47		12
15.52	29	209	212	211	223	47		12	870
15.56	33											Shut Down
2/12/2024   BOTTOM COWL OFF TOP COWL ON   OAT 24C
9.45	0	109	130	115				1
9.49	4	141	160	155
9.52	7	171	176	176			ON	6
9.56	11	177	181	181	178			7
10.00	15	185	189	189	186	48		7	870
10.04	19	192	196	196	198	47		8	870
10.07	22	196	200	200	204	47		9
10.10	25	196	199	200	208	46		9		12.5
10.13	28	198	201	202	211	46		9			40/4
10.15	30	200	203	203	212			9
10.18	33	202	206	206	215			10
10.21	36	204	208	208	217			10		12.4
10.24	39	203	207	207	219	45		10		12.2	38/4
10.25	40											Shut Down
2/12/2024   BOTTOM COWL ON TOP COWL ON   OAT 31C 3 Fans flush with areas in-between blocked off
13.54	0	94	104	111	109			0	860	14.1/13.0	30
13.57	3	143	162	154	135			1.2
13.59	5	173	177	177	153		ON	7	860
14.02	8	189	193	194	180			9	850
14.05	11	198	202	203	194			11
14.08	14	209	213	214	209			12	900
14.12	18	219	223	224	225			14	830
14.14	20	222	226	227	231			15			brief taxi @ higher rpm
14.16	22	226	230	231	236			16	820
14.19	25	227	231	231	244			16		12.7	34
14.20	26	228	232	252	245			16			17psi = 1.2BAR
14.22	28	226	230	241	241			19	Shut Down	20psi = 1.4BAR (max pressure)
24/12/2024   BOTH COWLS ON 3D Prototpe1 Rear Shrouds 4 fans two large two small  BAD DATA (Melt down)
15.56	0	77	87	77				0	NOTE Fans melted, shrouds melted BAD DATA
16.01	5	167	171	172	147			5
16.05	9	175	182	182	177			6
16.08	12	189	196	196	190			6
16.11	15	199	206	206	203			6	900	13.4	37
16.14	17	209	216	217	216			9
16.17	20	217	224	224	227			10	Rear shrouds shifted out of position
16.22	25	233	226	233	243			12
16.26	27	239	232	240	247			13
16.28	29	236	243	243	251			14
16.30	31	238	245	245	254			14				Coolant boiled
27/12/2024 BOTH COWLS ON No fans or Shrouds – BASE LINE NUMBERS  Both pumps on after 110F										OAT 23
13.59	0
14.01	2	124	128	128	103			3	880	13.9	47/1
14.04	5	149	152	152	127			6	910	13.7	46/2
14.06	7	167	171	172	152			8	920	13.5	43/2
14.08	9	177	180	181	165			9	870		42/2
14.10	11	187	186	190	177			11	900	13	39/2
14.12	13	194	198	199	190			12	910	12.8	38/2
14.14	15	202	206	207	201			14	910	13	37/3
14.16	17	209	213	213	210			15	930	12.9	36/3
14.18	19	214	219	220	220			17	890
14.19	21	STOP
29/12/24 BOTH COWLS ON 4 Fans (two large 2 small) Good new shrouds							OAT 28				Pump on at 104F
15.32	0	85	100	98	85				1010		39/0
15.36	4	139	143	143	122				860		46/0	aux pump on
15.38	6	155	159	160	140				910		44/1	(no change)
15.40	8	167	172	172	156				910		43/3
15.42	10	176	180	181	172				910		41/4
15.44	12	184	188	189	184			9	920		40/5
15.46	14	192	197	197	195			10	920		38/6
15.48	16	199	204	204	205			12	920		37/6
15.50	18	206	211	216	214			13	930		36/7
15.52	20	211	216	221	222			14	940		35/7
15.54	22	214	212	226	229			15	950	12.7	34/7
15.56	24	218	227	232	233			15	850		34/3
15.58	26	224	232	231	245			16			35/3
15.60	28	226	239	239	242			21			36/3
											Pump on at 104F
31/12/24 TOP COWL ON BOTTOM DUMMY COWL WITH COWL FLAP HOLES 4 Fans (2 large,2 small) SLIGHTLY MELTED new shrouds												OAT 27C
15.55	0	89	103	98	60					14.1	28/1
15.59	4	137	103	142	123			4			44/1
16.01	6	153	157	157	143			5			50/1
16.03	8	161	167	167	159		AUX OFF	6			43/0
16.05	10	169	175	175	169			6	850		43/0
15.07	12	176	182	182	178			7			42/0
16.09	14	180	187	187	186			8			41/1
16.11	16	186	192	192	193			8			40/1
16.13	18	189	196	196	200			9			40/1
16.15	20	193	199	199	205			10		13.3	40/1
16.17	22	194	200	201	209			10		13.4	39/1
16.19	24	197	204	203	213			10		13.3	38/1
16.21	26	200	206	206	217			11			38/1
16.23	28	202	209	209	219			11		13.1	37/1
16.25	30	205	212	212	222			12			37/1
16.27	32	208	214	214	225			13			37/1
16.30	35	211	215	215	230			13		12.8	34/2
1/1/25 TOP COWL ON BOTTOM DUMMY COWL WITH COWL FLAP HOLES NO FANS								OAT 23C		Pump on at 104F
14.50	0	75	74	75	71
14.53	3	124	128	127	101			2		14.1	30/1
14.57	7	147	152	151	123			4		14.1	29/1
15.59	9	163	169	168	142			5		14.1	28/1
15.01	11	177	181	181	159			6		14.1	28/1
15.03	13	189	193	193	172			8		14.1	28/1
15.05	15	200	205	204	185			9		14.1	28/1
15.07	17	208	213	212	196			11		14.1	27/0
15.09	19	213	217	217	209			12		14.1	27/0
15.10	20	221	224	224	213			14		14.0	0/0
				SHUT DOWN		temps too high
17/1/25 BOTH DUMMY COWLS, TOP & BOTTOM FLAPS, 4 FANS						OAT 21	Pump on at 104F
9.11	0	77	84	85	67		AUX ON
9.14	3	113	115	115	81			3	800		45
9.16	5	132	138	137	108			4	820		44
9.18	7	140	143	142	121			4	830		43/2
9.20	9	147	150	150	132			5	830		42/2
9.22	11	152	155	156	145			5	840		42/2
9.24	13	156	159	159	154			6	840		41/3
9.26	15	160	160	164	164			7	850	12.7	41/3
9.28	17	165	169	169	169			7	850	12.7	40/3
9.30	19	168	173	172	175			7	860	12.7	40/3
9.32	21	170	174	174	180			8	860	12.6	40/3
9.34	23	174	178	178	184			8	860	12.6	40/3
9.36	25	181	185	185	189			9	860	12.6	39/3
9.38	27	184	189	187	193			9	860	12.6	39/3
9.40	29	187	191	190	197			9	860	12.6	39/3
9.42	31	187	191	191	200			9	860	12.6	39/3
9.44	33	189	193	192	202			9	860	12.5	38/4
9.46	35	189	193	193	204			10	850	12.5	38/4
9.48	37	192	195	195	206			10	850	12.5	38/4
9.50	39	191	196	195	207			10	850	12.5	38/3
9.52	41	192	196	196	209			10	860	12.5	38/3
9.55	44	189	193	193	210			9	860	12.4	37/3
18/1/25 BOTH DUMMY COWLS, TOP & BOTTOM FLAPS, NO FANS (removed)							OAT 24C	Pump on at 104F
10.09	0	77	78	79	66		AUX ON	0	820	13.8	34
10.12	3	119	118	118	89			2	820	13.9	38
10.14	5	128	130	130	101			2	820	13.9	38/1
10.16	7	144	146	146	120			4	860	13.9	38
10.18	9	154	156	156	124			5	860	13.8	38/1
10.20	11	162	164	164	148			6	860	13.8	38
10.22	13	168	171	171	158			7	860	13.8	38/1
10.24	15	172	175	176	167			7	820	13.8	38
10.26	17	175	177	178	174			8	820	13.7	39
10.28	19	178	180	180	179			8	820	13.7	39
10.30	21	179	182	182	183			8	830	13.7	39
10.32	23	180	183	183	187			8	840	13.7	39
10.34	25	182	185	185	190			9	840	13.7	39
10.36	27	189	191	191	193			10	830	13.7	39
10.38	29	191	193	194	197			10	850	13.7	39
10.40	31	190	192	192	199			10	830	13.7	39
10.42	33	192	195	195	201			11	840	13.7	39
10.44	35	190	193	193	202			11	830	13.7	39
10.46	37	193	196	196	204			11	840	13.7	39
10.48	39	192	195	195	205			11	840	13.7	39

 

This last graph shows that the addition of fans was a very minor improvement and is certainly not worth the trouble.
They will block air in flight and use too many amps for the alternator to keep up at low RPM.
The result with the huge cowl flaps is not far off both bottom and top cowls removed, a good result.
The cowl flaps may also be useful in flight for prolonged climbs or just after take off in hot conditions.

All this testing work has put me in a position to confidently cut up my good white cowls.  Let me know if you need a very expensive but now useless set of two ply Long-EZ cowls! Despite the carbon they are really flimsy.

 

 

 

Date:  01-16-2025

Number of Hours:  16

Manual Reference:  no ref

Having solved the basic concept of how to open and close these flaps. It was a matter of refining the process.

I made a bunch of fittings although something tells me these will not be the final ones. In a wide eyed way, I had hoped so.

I got the arms working after a while with aluminium pieces and offsets with spacer inserts. A problem is that because of the compound curves the arms done pull straight up and they need to be in a bit of an arc. My little 9V battery didn’t have the power to pull them through.

Add a 12v aircraft battery and its a monster.

I refined it a bit but this is not a final result. After some thought I ordered some incorrect parts and have just ordered some more that will work. I’ve also lined up a machinist to make insets. I’ll show all this in a later post, assuming it works, or even if it doesn’t. In short I’ll use rod ends that are adjustable to replace the piece that bolts to the cowl door in the picture.

Meanwhile I have the proof of concept so lets move on.

I marked out on my dummy top cowl the biggest doors I could make fit.

Given its a practice piece, I cut the holes out in about 5 minutes, freehand. Looking in from above, the pipes are exposed top and bottom. This is the source of the under cowl heat so if its visually open, the hot stuff will get out.

These are huge flap openings, go big or go home right? It will be some work to ensure the final cowl work is very strong with so much structure removed.

I added the same offset hinges to the other three flaps and secured them with screws.  With the one bottom flap with a test mechanism that’s one flap open and taken care of.

Using the fittings I’d made earlier I secured the others with scrap aluminium. All I need is something that stays open and strong enough to go to the runup bay for testing and back again.

This has to be the money shot. An F16 look? I am ready to test.

 

 

 

 

 

Date:  01-11-2025

Number of Hours:  25

Manual Reference:  no ref

The 11 days since my last post have been busy.  The data tells me I need to vent the bottom cowl AND the top cowl. So first up I need a new top cowl to do my testing with.

I covered the top cowl in duct tape as usual, trying to get it as smooth as possible. Not that it matters except to work on my build standards. The resultant dummy cowl will be stored in my roof space when all the testing is done. I don’t think I could throw it away… yet.

This time I used up all my PVA release fluid. We will see if it makes the cowl easier to get off.

Despite the so called slow cure West epoxy hardener it was a real race against time to get the two plies of cloth wetted out and pushed and coaxed into place.

I spent well over five hours trying to get this cowl off the mold. In the end I had to make two cuts in the top, then it sort of popped in a couple of places. I quickly repaired the two cuts and after a trim we have a dummy top cowl to play with.

While it still needs more fitting work it is going to be fine for testing. I may have to reinforce the end as I did on the real top cowl so it doesn’t impact the spinner. We will see how it it after I get a few fasteners in place.

The cut-out for the flaps I felt was not quite in the optimum position relative to the exhaust pipes, the main source of heat. Here I’ve marked where they should go.

So I repaired or ‘put back in’ the flap doors!

After re-cutting the doors I am much happier with the position now. I moved this one back a little to leave room for the radiator fan. It was a little close.

Previously this door was further to the left, now it is well clear of the outer exhaust pipe. I will alter the closeness to the edge of the cowl when I do the real cut-outs. For now what I have is OK for testing.

You can see good visibility for the pipes now. Which suggests the hot air will get a better path and so improve the ground cooling when the plane is not moving.

Yes its not pretty but I spent a while getting the offset hinge working. It has to clear the cowl as it opens downward. Its the same type of hinge I used on the oil door. I have reverse the spring but its not strong enough to close the door on its own.

Despite the hot glue and ice cream sticks, it works quite well.

I’ve bought this mini 2″ push length linear actuator. Yes I know they don’t like the heat. A cable is out of the question, no room. So somehow I need to make this work and then shield it from the heat.

I worked non stop for several days on this problem. I got the actuator to work ‘a bit’ with partly opening the flap door but it was pretty awful. I don’t have room to make this actuator upright and it would need a lot of mounting if I could. Even then it would get knocked over sooner or later when I pulled the cowls.

I tried every configuration that I could think of. I went to several engineering shops expecting a solution. I just wanted a solenoid type until that would move an arm 180 degrees. I thought I could make that work. Devices were either way too heavy, thousands of dollars each or both.

I went down the remote control cars and planes route, the big ones. They have quality servos. I spoke to several experts. Wrong voltage and they need a controller, it was just getting too complex when all I wanted was a thing with two wires that I could reverse… like the actuator! …but I couldn’t seem to make that actuator work with the space issues and they just didn’t seem to support the door very well.

I found a nice cowl flap mechanism for RV planes. It was around $800 and out of stock and wouldn’t fit anyway. I did like the idea…so I started to mock up something quite different to that but ‘inspired’ by this stock item.

Does this give a bit of an idea?

Here’s the basic movement. This is a long way from useable but I was onto something. I had to put the bottom cowl back on and work out clearances of course and then where to mount it.

It tucks in here. I did have to move the exhaust pipe clamp but a good days work will make new supports if I need to. I might have to move a hose as well. We will see.

It will fit on this opposite side but I’m right near the exhaust pipes. I think I can build a shroud for them and it might survive will lots of shielding.

I’m getting a good opening. It took a few hours to work out the gearing and several ice cream sticks.

Here is the proof of concept. It works. I have a cowl flap mechanism. Now to make the parts a little more carefully!

 

 

 

 

 

 

 

Date:  12-29-2024

Number of Hours:  3

Manual Reference:  no ref

I used a lot of gaffer tape and spare heat shield material to secure the fans and shrouds for the next test.

You can see the fans really tightly in place. Its just a quick test and gaffer tape works on everything. I think they have some on the ISS (International Space Station).

Here’s a bit of a walkaround before I headed off to the runup bay for testing.

It turns out that gaffer doesn’t work that well when things get a bit hot. Goodbye expensive fan. 🙁

In fact I should have pulled the pin earlier.

After I shut down the engine and coolant pressure continued to rise. At 18psi it will blow the cap, or rather the bypass valve in the radiator cap with operate to send coolant from the expansion tank into the overflow bottle. I saw a bigger number than 18psi in the data.

If its still too hot, that overflow bottle will fill up and as a last resort it has a little hole in the cap.

My expensive 3D shrouds are not what they used to be. See item #6

A few lessons learnt.
1. If the test is showing high temperatures, just abort early.
2. Minimise the under cowl testing temperatures, I might be stressing wiring and other components.
3. Reduce the time in the idle, I REALLY don’t want to glaze the cylinders.
4. I need to reduce all this taxi time, so only necessary tests and finger crossed about the glazing.
5. I need to do better than gaffer for securing heavy fans onto radiators in high heat conditions.
6. 3D printed parts with cheap plastic are NOT heat resistant.

Ahh but I won’t be put off easily. Here we have another set of shrouds again nicely printed. The data from the last test was not useful because of the shroud failure. I want to know do these fans and heat shrouds work??

This time I got excited with foil tape, lots of heat shielding, cable ties and even lockwire.

Here’s a bit of a look around. Off I went to the runup bay again full of hope and good cheer.

The fans did sort of stay in place but the shrouds were again heat ‘effected’. I could use them again but its not a 100% seal. At least I got some data.

Yes my chart skills have improved. You can see Both Cowls on with melted shrouds and Both Cowls no fans is pretty much the same temperature. I want to be around 200F for take off so I have a bit of margin when the aircraft goes from idle rpm to full power. Its going to get hot, fast. I will test this in the runup bay and/or on the runway before first flight. I don’t want a boil over after two minutes!

The numbers we want to compare at ‘Both on no fans no shrouds” (orange) with “Both Cowls Good fans/Shrouds” (light bluish) The fans and shrouds give me only about 10F improvement and maybe 5 minutes before we get to the 200F at the 15 minute mark.

This tells me the whole fan shroud thing is doing very little really and given I want twice that time before takeoff and still be around 200F, I am nowhere near my design goal.

CONCLUSION: I need cowl flaps on the bottom cowl and maybe on the top too. I do want to confirm after getting the bottom cowl with cutouts for flaps tested with and then without the fans. With hot air escaping I would then find out if the whole fan/shroud thing is worth the trouble. I suspect not, but data is king.

CONCLUSION UPDATE: But wait, its 1/1/25 and I did the test with the cowl cut outs and NO fans. I was completely surprised. The fans, now that there is a path for the hot air to escape do a huge amount of work unlike before the opening. At just under 20 minutes with fans 189F and without fans 221F and rising fast! That’s 32F difference. The oil was 200 vs 213 without fans again at only 20 minutes.

I did remove the fans for this last test and they are looking quite melted and sad. I am going to have to do the next set in Carbon fiber as a hand layup, the plastic is a big fail. I’ll get more of these printed in something I can use for a mold so at least it will give me a shape to work with.

It looks like I am going to need those fans and shrouds after all which is disappointing. The plan is to now go ahead with a dummy top cowl and do similar cut outs. I’ll have to make doors for them of course and get that all working in the mock up. I will then test again, with and without fans and shrouds and make a final decision. At this stage it looks like the fans are back in the mix again unfortunately. The doors and going to be very tricky to get working so that’s my January 2025 already planned out.

 

 

 

 

 

 

 

 

 

 

 

Date:  12-17-2024

Number of Hours:  9

Manual Reference:  no ref

While working towards the next ground cooling test in the runup bay I began a side project. If this fan and shroud thing doesn’t work then I’d better have a Plan B. That would be openings in the cowls leading to cowl flaps. My testing said bottom cowl off, top on was pretty good for cooling. So I need to try openings in the bottom cowl right?

I don’t want to cut up my good cowl just yet so I need a dummy to work on that I can throw away later. A test piece.

Those two bits of carbon were around $200 and then I need a lot of epoxy. It was a meter of cloth per layer and that was barely enough. I think three ply would have been better but this is going to just be for testing and only needs to hold up to taxi out to the runup bay and back.

Of course my ‘mold’ or plug was already made. I just covered the original lower cowl in duct tape. I did use some wax on top as well. I will use a lot more next time or hair spay or even proper mold release.

Yes not pretty. The carbon was a little wetter than usual. Remember its a throwaway.

I managed to cover the whole cowl and gave it a day to cure..

It was VERY hard to get this dummy off the mold, probably because of the compound shape. It looks a bit better on this side.

Here we are off the mold. I did a bit of pour foam on the boat tail later to stiffen it up. Overall I have what I wanted.

The ‘dummy’ lower cowl does fit well enough.

Here we are a couple of weeks later. I’ve done several taxi test runs keeping good temperature data. I am going to need cowl flaps, no question. I have cut out my ‘flaps’ and I will report on the success or otherwise in terms of cooling. If it goes well, then making flaps that open and close will be quite a challenge. Its a good thing I have this mock up to try stuff out with.

Left side, note how I’ve gone right where those hot exhaust pipes are.

Right side, the same size cut out. Even with the pipes being in different places with the cylinder offsets I wanted them the same size and place. If I need them in flight I want the drag from the doors to be equal both sides.

It doesn’t look too bad, of course I am going to have to have them close in flight. This opening only setup is firstly to test the concept. I did do a test run Dec 31, 2024 but a LOT of things have gone on that I need to tell you about first.

 

 

 

Date:  12-24-2024

Number of Hours:  3

Manual Reference:  no ref

To solve this ground cooling issue I first need to make the fans more effective by building shrouds. That’s like a faring that sits an inch or so from the rear face of the radiator to gather air from the entire back area to be pulled through by the fans.

First I need a prototype for a proof of concept. I could just mock something up which would take quite some time. Then it it wasn’t quite right I’d have to start from scratch and build another, and another…

Wouldn’t it be great if someone could wave a magic wand over the radiator instead? After that they might made a gesture and point to a magic box. After more time a perfect shroud would just appear!

Here is a magic wand with a modern day wizard getting the 3D image.

We need the cowl too so the images can be stitched together to give us the clearances. Wands are boring grey oblong boxes these days. It is still magic.

Here we have a problem revealed. A third fan is just too close to the induction pipe. In fact it will not fit. We could make that fan ‘remote’ but for now we will make a two fan solution and test with that.

This is a magic box. It is used after the magic wand.

Magic is slow.

Magic is really slow. Layer by layer things pop into existence. This is just showing off.

The first half of the shroud has now appeared and the fan fits perfectly on top. The second half will be ‘printed’ and then I’ll install it ready for testing. Of course the right hand side shroud also needs to be printed with this slow magic method.

With the two half printed we did some more testing. Is a bigger fan but offset better than a smaller fan not offset and a bit further back. I expected the bigger fan to have more ‘suck’ where we placed a bit of paper at the respective openings. To my surprise the smaller fan sucked a LOT better.

OK a bigger fan centred would be better but I have space constraints. I have ordered another 4″ fan and we will be able, because of its size, space it a bit further back and in the centre in the next prototype.  This creates a ducted fan effect which also helps.

I was in for another surprise. I had asked ‘Spal’ fan support how will I go with wear and tear when the fans are freewheeling at high speed. I didn’t get a good answer but he pointed out that as the incoming airflow gets stronger the fan motor works less. I thought it might fight against the flow and try to slow down. Wrong, it requires less and less current. Until it needs zero, then it acts as a generator and makes current for my electrical system!

My avionics guy knows all about this and showed me with meters that my fans freewheeling produce current so my alternator doesn’t have to work as hard. It should also help to accelerate the air after the fans and so reduce cooling drag. I’m liking these fans a bit more.

Here we have our prototype version 1 shroud. This is for proof of concept. I’ll need the second shroud printed before I can gaffer tape them on and do some testing.

The RHS shroud has been printed. We ran out of black ‘thread’ so white it is. Not so heat resistant it turned out later.

You can see a nice offset on the smaller fans. Going back a little further gives more ‘suck’. We are as usual very limited by the room available.  The dip in the middle is for an exhaust pipe. Did I mention the lack of space?

This is how it sits with gaffer tape. I did add some heat shielding to protect the plastic.

After taping the ducts in place a little preliminary testing was in order.

Here we have proof of concept. Before the shrouds there was very little ‘suck’ on the radiator face side. Barely enough to hold a sheet of paper. Now we have something worth testing. The next step is back to the runup bay and see what I have and collect the data. I am or rather was, hopeful of a good result.