Jun 24, 2001, 10:56 PM
Registered User

Yellow AC F-4 twin

Inspired by the boldness of the recent post about the conversion efforts by Plane Crazy of a Yellow AC F-18 and the belief that the technology has finally arrived to allow this conversion to be successful, I now find myself head long into a Yellow AC F-4 slimer to twin EDF project.

It wasn't really planed that way as I had bought the Phantom used, but never flown. This is very fortunate for the Phantom as I am sure it would have met it's demise on the first flight. The whole project was a mess and I spent a lot of last fall stripping it back down to the bare fiberglass, had already reworked the glass on the wings, and everything was back to the primered stage. It had set since September as I had lost interest in the project after discovering EDF's. Lost interest that is, until I started reading about the Yellow AC F-18 conversion. Well the Phantom and the F-18 specs aren't that different. A couple of emails back and forth with Gordon (Plane Crazy) and by 2:00 AM the next morning, the old duct work and formers were removed and plans were underway for a F-4 EDF conversion.

Well, my time is currently very limited and progress will be slow, but the F-4 will get a retro fit. The Phantom actually worked out very well as far a available space and dimensions. From the intake duct to the exhaust is 36". The rear of hatch opening on the Yellow AC version of the plane is right at 24" from front of the duct. This means I can place the fans just inside the hatch where they are accessible and will also be right at the 2/3 intake 1/3 exhaust placement that Gordon suggested.

The rear ducting is very straight forward as there is nothing between the fan and the exhaust ports. The real challenge is the front duct and this is where I am looking for advice. The intake on the Phantom is basically a rectangle 3.079 X 5.04 (15.5186 Sq In.) which needs to flow (morph) into a round duct with a 3.543 9.7818 Sq In). There is plenty of air coming in, I would just like some suggestions on how to design the duct to keep the air flowing properly through the duct. Do you get to the round duct ASAP or does duct change take place over the whole length of the duct. Im looking to some of you more experienced designers to help make this project a success. I have read lot of the old post on this matter and the design of this front duct seems to be most critical.

Thanks for your help,

Ken Rose

Sign up now
to remove ads between posts
Jun 25, 2001, 01:30 AM
Registered User

Shouldn't be that hard. Get a big chunk of bluefoam the length from intake to fan OD, tack glue a 1/32 plywood, 3.5 inch disk on one end and a 1/32 ply "intake shape" on the other and sand away! Actually if you can hot wire, a lot of the excess material can be removed without dust. If you need to laminate your foam blocks, use 3M 77. Now since dissolved out foam does not leave a very good surface finish, You'll need to apply some kind of finish to the foam (prior to glassing it). Suggest some lightweight spackle and water based polyurethane. When done, apply mold release to this "consumeable plug" ,then apply the appropriate amount of glass to the foam and then use mechanical methods to get the foam out.

I used one other way to make (3) identical Exhaust ducts for my XB-70 (which incidently transitioned from round to rectangular in 6 inches): made the male plug from balsa and then monokoted it for a quick mold release! After each part was cured, I slit the full length of the duct and peeled it off the plug. Sometimes the monokote was damaged in the removal process, oh well, more monokote! After the part was removed, I taped the seam back together and applied a 1 inch strip of 2 oz cloth to the seam and presto, finished duct.

Be aware that intakes operate at below atmospheric pressure and will collapse if too weak. An educated guess would be minimum 3 layers 2 oz cloth plus some attachment to the fuselage here and there. By the way, your intake size seems too good to be true! A rule of thumb is ~ 115 (?) % SFA for the intake and 90% for the exhaust. Sounds like you could accomodate bigger fans. On most scale jets, the scale intakes are way too small. Two other good rules of thumb are 80 (min, 100 better) watts-in/lb and maybe 35-38 oz/ft^2 maximum wing loading. Does it all fit?

Best of Luck to Ya,

Steve Manganelli, aka Prof. Maneuver
Jun 25, 2001, 08:58 AM
Registered User
Ed Waldrep's Avatar
The intake ramps/splitter plates on the F-4 were moveable, to vary the inlet airflow at various altitudes and airspeeds. You could partially close off the ramps to get a bit closer to swept fan area, but still stay above it. A lot of excess inlet area would produce more drag and limit top speed, but would have better static thrust than smaller inlets. How much more I can't say, but it seems reducing the inlet area a bit would produce a good balance.
Is this the 91 sized F-4 for a single or the bigger twin version? I seem to remember Yellow having a twin but I could be mistaken!
Jun 25, 2001, 08:59 AM
Electric Airplane Junkie
bhchan's Avatar
Originally posted by Prof. Maneuver:

Shouldn't be that hard. Get a big chunk of bluefoam the length from intake to fan OD, tack glue a 1/32 plywood, 3.5 inch disk on one end and a 1/32 ply "intake shape" on the other and sand away! Actually if you can hot wire, a lot of the excess material can be removed without dust. If you need to laminate your foam blocks, use 3M 77. Now since dissolved out foam does not leave a very good surface finish, You'll need to apply some kind of finish to the foam (prior to glassing it). Suggest some lightweight spackle and water based polyurethane. When done, apply mold release to this "consumeable plug" ,then apply the appropriate amount of glass to the foam and then use mechanical methods to get the foam out.

Steve Manganelli, aka Prof. Maneuver
I used this method to make airducts for race car(1:1 scale). After you achieved the shape of the duct you desired, simply tape the duct with packaging tape(or Zagi tape),make sure the duct is fully covered by tape, less wrinkles are better. Apply fiberglass over the taped duct. When the fiberglass is cured, pour solvent(acetone) to desolve the foam and remove the tape from the inside.

Brian, an EAJ.

Jun 25, 2001, 09:40 AM
Flying Welder Pilot
Plane Crazy's Avatar
Ken- I like Ed Waldrep's idea of making the intakes smaller to get closer to the fan swept area. It would also allow you to get rid of the sharp intake lips. If you closed down the intake on three sides by .561" your area would be about 9.839 sq in. Then you could round the intake lips and get a little better flow too. The intake rings on the Midi's are .475" thick, so you would be a little bit thicker.

Good luck,
Jun 25, 2001, 10:00 AM
EDF Head
Haldor's Avatar

This is what I did - foam plug sanded to shape then covered with thin 2" wide packing tape. Then applied cloth/epoxy resin and melted the foam out with acetone. Then pull out the tape and presto - done!
Jun 25, 2001, 10:07 PM
Registered User
Keithf's Avatar

I've just gone through the same process for a Grumman Cougar I'm building for 32 cells & WeMoTec HW730E. My inlets needed to start at a triangular shape & transition to the annular fan shape. To boot, each inlet curves in plan view and side view, and twists along its length. Inlet area is only 90% FSA.

I decided the only way to go about it was to be reasonably scientific, as duct losses were going to be costly with fully bifurcated ducts and low inlet area. So, I drew up the starting and ending cross sections, plus 2 known midpoint sections (long story) and proceeded to a friend's place with CAD software. After scanning the drawn sections, we morphed the intermediate sections, such that I had 18 cross-section templates transitioning from front to rear.

Then came the surprise, which may be of interest to you.

I had calculated how the cross sectional area of the duct was to transition from the 90% at inlet to 100% at fan, so we proceeded to check the areas of each progressive template, to see how they stacked up. You would expect that if you start with a section of known area and transition it to a similar but slightly different section of known area, then the area transition should be fairly uniform and consistent. Not so. Every section needed to be strectched or scaled to achieve the correct area.

As an example, one group of templates had to transition from 98.5% area to 100% area, being the last 15cm length of duct immediately before the fan. Sections were very similar. With each template being spaced just 3cm apart, the area should have increased by about 0.3% per section. One section saw an increase of over 5 percentage points. (to about 104%)

Each time the area changes, there will be a velocity or pressure change in the airflow. As we live on a planet where every law of physics is harnessed with less than 100% efficiency, there will be efficiency losses with every such velocity or pressure change.

So, if ultimate performance is not important, then by all means sand away with you chunk of blue foam. If you want to squeeze every last bit out of it though, you'll probably have to be a little more precise.

BTW, I also agree with the comments about reducing the inlet area. Given good duct and inlet lip design, inlet areas as low as 60 or 70% do not appear to create substantial losses. 90% is generally considered safe for a single duct, with good overall design. Given that your inlet lips are probaly going to be too sharp, based on the full size, then you may want to stay above 90%. A common rule of thumb appears to be that bad inlet lips can reduce static thrust by up to 30%, and require inlet size to be increased by up to 60% to compensate. Enjoy.


Jun 26, 2001, 02:41 AM
Registered User
You hit the nail on the head. I am concerned about area changes as the transition takes place. I did a trial (morph) using Corel Draw 8 and it produced a rather unique design, it did kind of a twist as it moved from the rectangle to the circle. The Phantom does have a very sharpe intake and I was wandering about starting with the full size opening and using a flare inside the duct, similar to that of the inside of a trumpet, to reduce the size quickly at first. Another thought was to split the duct and move part of the air over my batteries. This does produce drag, but would that be ofset by the benefit of cooler cells?

I just got in having got my firts server upgraded to Win2K and my brain is mush. I'll look over the posts again tomorrow, later today actually, when my brain has had a bit of rest.

Thanks for everyones input, I'm open to any ideas that come my way.

Jun 26, 2001, 11:18 AM
Flying Welder Pilot
Plane Crazy's Avatar
Ken- One trick to morphing with Corel Draw is to cut the circle so that it has the same number of end points as the rectangle. You can cut the circle at places that corespond with the points on the rectangle. This would give you a better transition between the shapes. It will be difficult to calculate the area of each section though.

I used a program called Solidworks (a solid modeler CAD program) to create the solid model and cross sections of my ducts on the F18. You can easily cut the solid model anywhere and select the section surface and it tells you instantly what the area is. I hate to tell you how many sections it took me to get the ducts around the main gear and the taileron pivots. Then I plotted out the sections and cut foam the thickness between each section and sanded it to match each template glued the whole mess together, sealed and smoothed it over with automotive clay.

I have spent more time on ductwork than anything else on the F 18 project. 3 trys so far. If I had a computer model of the Yellow F 18 things would have been easier.

Jun 26, 2001, 08:40 PM
Registered User
Keithf's Avatar

I think your first design question comes down to inlet lip shape. If you can't get a 2:1 elipse shape, with thickness of around 15% duct diameter (another can of worms) then you are probably going to have to increase inlet area above 90%. This should get reasonable static thrust performance, at the cost of high speed performance.

After establishing lip shape, inlet size can be established with a little more certainty. Then the transition.

The key here seems to be keeping everything smooth & gentle. Wall angles are not supposed to diverge at an angle greater than 4 degrees, although from my experience anything near to 4 degrees rapidly gives enourmous volume increases, so that shouldn't be a problem. Convex curves in particular should be kept as gentle as possible, to avoid airflow separation. Corners between duct walls should be radiused. A circular duct is the ideal shape, so you would probably want to transition from the rectangular inlet sooner rather than later.

As losses are magnified by increased airflow rates, airflow should be kept as slow as possible, as long as possible. If you are increasing the area (ie inlet area less than FSA) then make any increases immediately after the inlet, before losses from curves or shape changes are encountered. If you are decreasing the area, then leave it until immediately before the fan.

Hope this helps.
Jun 26, 2001, 08:47 PM
Registered User
Keithf's Avatar

One other thing. Research and empirical testing (not mine) suggests that the fan should be located as close to the exhaust as possible, rather than at some arbitrary point along the duct. Ideally, you would probably position the fans such that there is just enough length to install the exhaust cone immediately after the fan.

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion JHH F-4 Twin 90mm EDF Conversion PhilLin Electric Ducted Fan Jet Talk 427 Feb 16, 2017 06:05 PM
For Sale Yellow Aircraft F-4 Phantom II - NIB skyhawk Aircraft - Fuel - Airplanes (FS/W) 6 Aug 23, 2005 07:00 PM
Re: Any Yellow Aircraft F-18 Twin owner here? X-spec Fuel Jet Talk 10 Mar 16, 2005 08:19 AM
Yellow Aircraft F-4 RPM Fuel Jet Talk 16 May 11, 2004 03:33 PM
Cool Yellow Aircraft F-18 Twin EDF GregG Electric Ducted Fan Jet Talk 39 Jul 18, 2001 01:21 PM