P47 21" composite honeycomb Kevlar and Glass fabric - RC Groups
Thread Tools
Dec 16, 2012, 05:13 PM
Registered User
pirulin's Avatar
Build Log

P47 21" composite honeycomb Kevlar and Glass fabric

I do not write much on your forum, but I want to share this experience started a year and a half.
I therefore subject for those who want to build a parkflyer light and strong.
This construction technique is suitable for small powers. So for those who like the 'all plastic' let's go. Sorry for my English its'nt perfect. Good read.

I want to thank those who by their perseverance managed to develop engines, electronics and technicals for our little flying toys.
Especially so thank you vladimir88 for your work on brushless motors with gearbox with 1S powersuply

My initial goal was:
Fly a model airplane light, solid 1P 1S (3.7V) with electronics that I use for my micro-helicopter.
This is a difficult challenge, but playable, given my limited experience.

Final specifications are as follows:

1S1P lipo battery 1000Mah 15C 28 gr
Walkera RX2609 Receiver 2.4 Ghz modified 2,5 gr
WK03 digital servos 3,5 gr-2
Helicopter Ball joints from Walkera 4G3 or 4G6
Home made gearbox ratio 1/5
HP06 motor 3.7V. Kv 16300. I max 8A . 7gr
ESC XP12A V2 4.5 gr
Walkera 2801PRO 2.4GHz Radio
Propeller GWS 7X3, 5 // Trust at 4,2 V // 5.85A : 100%=170 gr at 60%=140gr


ELEV 0.8 mm Carbon rods
AILE steel rods 0.5 mm


Kevlar 36gr/m2
Glass 25 gr/m2
HM Unidirectional Carbon reinforcements for local
unidirectional kevlar
Honeycomb: 1.5mm 44 gr/m2
LY5052 resin and HY5052 Ceiba Geigy

Wingspan:54 cm (21 inch) Area 4,8 dm2 / 74.55 inch2
Length: 45 cm (17 inch)
AUW : includes battery , 150 gr. ~5,4 oz

Soon, I will explain how to get your bird without the ghost marks of the honeycomb despite the delicacy of the skin that are ...

Last edited by pirulin; Dec 20, 2012 at 07:36 AM. Reason: trust, wing area, complete weight and picture added
Sign up now
to remove ads between posts
Dec 17, 2012, 06:25 PM
Registered User
pirulin's Avatar
To start, I made a master mold medium covered with several layers of epoxy resin. Finally after sanding it provides a very hard surface especially the resin penetrates into the medium, the finish of the epoxy coating is performed with water paper grade 600. The surface layer is sprayed with many coats it takes to get a deep gloss and gloss finish paper with water grades 600, 800,1000 then polishing with polishing compound for composite molds. I don’t develop well on the subject because there are a lot of tutorials on making masters. 5 times polishing with wax partall without siliconne.
The master is ready to be placed on the joint plane and achieving the epoxy half mold.

The edges of the joint plane are covered with a PVC sheet plated 0.2 mm around the master and chewed with plasticine in the corners. The centering studs are made by molding PVC vacuum on glass beads. Each form obtained is cut so as to have a half sphere. These blocks are then filled with plasticine and glued around the perimeter of the first half molds.

The mold is laminated fiberglass with a weight of 50 g/m2 to increases 300gr/m2. I put a wood reinforcement to mold the fuselage.
So far nothing very complicated. I left half molds sit a few days before proceeding to the stratification of parts P47.

The statification:

I chose to use Kevlar and not carbon for two reasons: firstly because the kevlar is less sensitive to shock than carbon, and secondly because I fear the influence of any carbon structure on the scope or range of the radio.

Kevlar is a taffeta 36gr/m2 the thickness of the laminate layer is vacuum of 0.07 mm.
The glass fabric 25g / m2 gives a layered vacuum thickness of 0.027 mm

A follow
Last edited by pirulin; Dec 17, 2012 at 08:28 PM. Reason: soory for my languages mistakes
Dec 18, 2012, 09:51 AM
Registered User
pirulin's Avatar
I did some tests under high vacuum to validate this technique before laminating the final parts to not spoil material and waste my time. I used two vacuum pumps: a single stage system (with a home made vacuum regulation to nearly 50 millibars accuracy ) and another industrial double stage the two vacuum pumps are fitted with non-return valves.
I tested also the flexibility and deformability of my vacuum bags, using successively trash bags Polypropylene 18μm, 22μm, 50 microns, it's cheap and it works well for the wings, for the fuselage itsnít suitable because the shape is very deep and complex to wed, so I used a commercial product. The perimeter seal is a seal butylene base, used by carpenters for roofing, cheap (3.84 euros / 5m) very effective.
Itís very important to have a perfect seal, otherwise you could not sleep during the polymerization, it has happened to me, the result was successful, but 16 hours including 8 hours laminating parts with monitoring the bag which leaking vacuum regularly.
You must be certain about your hardware before making the parts you want to achieve.
The preparation of all stages is essential. The use of plasticine mold can helps the fabric to wed the tight spots of mold: Karman fittings, angles etc. .. To pump the excess resin, I use cellulose wadding (toilet paper), the perforated film used for the drainage of the resin is a commercial product.
Principle: compressing the impregnated fabric against the walls of the mold is evacuated and the excess resin by absorption and drainage device disposed on the tissue.

Take care with epoxy : is a dangerous chemical, wear gloves and goggles and if possible protect your skin with a barrier coat anti allergic.

First layer:
The mold is ready, he received the mold release wax + polyvinyl alcohol (PVA) to obtain a smoothly release and without mechanical stress. Glue the seal for the vacuum bag and pre cut the vacuum bag and the apertured film and the absorbent.
Before cutting the fabric, you can orient the fabric in the direction you think is best for optimal mechanical performances, you can also drag your tissues at 45 į to increase its malleability to fit the best possible mold. But be careful in doing this you will also increase its density.
n the case of Kevlar 36 gr/m2 once it has been slipped to 45 į or more, nearly double the density per unit area because the fibers become very tight! .. The weight of the finished piece will be penalized.
Cut the fabric according to a cardboard cutout template that you have previously made. It is better to put the kevlar fabric in an oven between 100 and 120 į C to remove moisture during Ĺ hour. While the fabric is in the oven, prepare your resin in small quantities, carefully weigh and measure the components very precisely (to the nearest milligram). Mix well, then degas the resin in the pot by placing under high vacuum.
You can now stratify the first layer in the mold with flexible bristle brush in order to donít injure the fabrics. With very fine fabrics such as those used for this construction, we must be also very careful when handling them because they are extremely fragile, if the frame is damaged, I urge you to wear latex or nitrile ... Adjust the part of tissue which extends beyond the cavity of the mold by protruding 5 to 10mm. Install the perforated film, taking care to have as flat as possible, even if the incision with scissors in the rounded or angular, then install the absorbent adjusting the same way and finally put the vacuum bag on the joint, making sure the seal. Install the cord that you can see from the pictures, it can distribute the vacuum evenly, and, protect the ends of the suction pipe so as not to injure the vacuum bag. Apply vacuum gradually phasing leaks or closing a valve vent free pre-installed for this operation, it allows you to take the time to properly positionnning your vacuum bag and adjust the layers under the vacuum bag and remove wrinkles when you make the vacuum. Then apply a vacuum as deep as possible.
In my case the resin cures in 24 hours, then it is completely stabilized after fifteen days. Before proceeding to the second layer it is important that the first layer is completely stabilized thatís the only way to avoid honeycomb ghost appearance on the skin.
Internal stresses of the first layer related to the curing of the latter will be completely eliminated only after the stabilization process. So after this first step, you can continue to work: the first layer is sanded lightly with a dry paper grade 80-120. The dust is sucked into the mold. The surface is possibly degreased with acetone if you work without gloves. Now we are going to ask the second layer that will interface between the tissue and the kevlar honeycomb.

To be followedÖ.
Dec 18, 2012, 03:38 PM
Registered User
pirulin's Avatar
I forgot to mention that for each layer laminate you can make a curing under vacuum in an oven, youíll get a more solid part, depending on the resin used, make sure in this case what is the process to be followed in levels of temperature and curing times thatís specified by the resin manufacturer, if these conditions are not followed, it can seriously harm the mechanical characteristics.

For my part I felt that, given the size of the final parts that eliminates the need for not proceeding with oven curing. The items I got, are already super crashproof.

You can see on the last picture of the two half wing cured, sanded and cleaned. The upper part is partially unmolded (it appears more clear ). In this case if it happened that the part demolds to the edge, a little skin of PVA layer can be applied to re seal the mold et the part against the resin, that prevents for the next step that resin can could pass under the previous layer and the mold surface.
Last edited by pirulin; Dec 18, 2012 at 03:44 PM. Reason: Comment forgotten for pictures : added
Dec 18, 2012, 05:47 PM
Registered User
pirulin's Avatar
Second step :

Previously you need to prepare your fabric as explained before and also cut the honeycomb as showed on the pictures. For the wings and tail I prepare also a thick PVC sheet to finally cover the honeycomb and so protect the vacuum bag. To save time, for the fuse, i donít use a PVC sheet on the honeycomb everywhere, but only on the edges. The new seal is glued all on the perimeter of the mold beach.
The first layer is sanded and clean, if you look at the kevlar itís injured by light sanding and many wire appears free, that is a good basis for physical grip for the next lay.
Now its time to glue the second layer and the nomex honeycomb core. The 25gr glass is laminated with a very low ratio of resin, it must just becoming transparent. Then I pump the resin in exceed with a toilet paper to let the glass surface very dry.
Then I install he honeycomb and temporarily I fix it with small weights regularly distributed. Now cover PVC sheet with pre cut to the exact size of the contour of the honeycomb in difficult places (corners etc. .. ) I stick the plasticine or on the entire surface of the wing to protect the bag against injuries caused by the honey. I puts temporarily my weights time to start the vacuum pump and cover the mold with vacuum bag. I take the weight remained under the bag holding the honeycomb and PVC sheet. I put now the weights on the vacuum bag, now I check that everything is in the right place before sealing the bag, I always leave a vacuum leak to gradually adjust anything that needs it. When all is right I aply a wide vacuum and let cure that, for 24 hours.
I'll post more after the new year because of my hollydays, i left my home.
Iwish you a good year end and christmas season
Last edited by pirulin; Dec 18, 2012 at 06:21 PM. Reason: another mistake due to my bad english, sorry
Dec 19, 2012, 05:17 PM
Registered User
looks like a nice build. did you go from initial goal to final specs or was there some feasibility study not included in your first post? maybe a simple power to weight ratio and then wing loading so we can follow your build. the plane looks very nice.
Dec 20, 2012, 07:07 AM
Registered User
pirulin's Avatar
Originally Posted by paulatgis
looks like a nice build. did you go from initial goal to final specs or was there some feasibility study not included in your first post? maybe a simple power to weight ratio and then wing loading so we can follow your build. the plane looks very nice.
Yes, you're right, it lacks datas, I'll complete the first post to be more understandable. Above all, I did an evaluation (weight estimate) depending on the area implemented to achieve my goal. I I'll show you that a little later because I go on vacation for a week and I lack the time to prepare a post
Again excuse my English.

Jan 04, 2013, 07:09 PM
Registered User
pirulin's Avatar
Happy New Year to all

Third and final step for the sandwich: Laminating the inner layer of the sandwich.

This method avoids having to make a positive mold to stratify the inner skin on the honeycomb, it is more economical, simple, disposable or reusable as precautions.

For the wings:

We'll have to sand the honeycomb, before that, I measure the thickness along the length of each half wing to determine the value of the internal thickness of each half wing, in my case it allows me to know zone to sand honeycomb as Iím be able to close the mold for final bonding of the bottom surface and the top surface. The edges of the honeycomb should be left not right but be sanded like a whistle to end to zero thickness, perimeter zones of the wing (leading and trailing edges) are also sanded to obtain a good grip of the last layer of the sandwich. Once done, I aspire dust and cleaning the mold, the mold beaches are re-treated if necessary with PVA. The sealant is placed on the perimeter of the mold, the mold is finally ready to be used for this last step.
I carefully prepare the fabrics by precutting using templates, I prepare the drainage film by cutting as the same size as the fabric, I also prepare the PVC sheet 20μm according to the same size that the template I used for cutting the honeycomb, I prepare the absorbent paper for cutting the fabric leaving one centimeter in exceed.
As you can see, I decided to mold a part of the fuselage with the bottom of the wing, it does not facilitate the work, but it is lighter than a piece glued to obtain this result. So we'll have to use cunning... this is where, things get complicated ...with my poor english!!!
My PVC sheets are cut as I said before, now we have to do that must perfectly matches the honeycomb, for that I use plasticine. So I put my sheets PVC on honeycomb, for the central section , I proceed piece after piece with PVC because the central form of the central bottom surface shape isnít developable. The central section composed by pieces of PVC sheet are linked with 20μm PVC tape . I do my best to shape each section of PVC and apply the tape to fix to each other, once this is done, I leave the form of PVC in place with weights and I plate plasticine to fill the hollow of this part of wing. The obtained shape would alone now. I return the shape to apply some PVC tape (on the central part only) is almost ready, it remains to drill the entire surface with hundreds of 0,3mm holes spaced between 8 to 10 mm, to drain the excess resin.

Now everything is ready: I install a newspaper on my worktop and I fix it with tape on the perimeter, I put the perforated film on the paper, then apply the 25gr/m2 fabric carefully as possible without fold. I apply the resin without excess, just to wet the fabric, it will stick by capillary perforated film, there must be careful to ensure that the fabric is well plated on perforated film without wrinkles! It is ready for installation of this layer on the honeycomb to obtain a beautiful sandwich.
I raise all perforated film and glass fabrics by catching the film perforated, be carefull, with very thin layers as used for this model, is very fragile and only asks to offend! Now I apply this layer hand taking all the tightest possible while I applied to the honeycomb and avoiding wrinkles forming on the honeycomb. Once the honeycomb covered with glass fabric and the sheet of perforated film, I pull gently on the edges to smooth the envelope as much as possible and I plate perimeter with my brush with soft bristles and a bit of resin. Now I apply the PVC sheet with plasticine, I have regularly to keep weight off, it allows me to chase bubbles trapped between the drainage film and perforated PVC sheet, if necessary I fixed the PVC sheet with tape. Once the unit is perfectly positioned so I can put the paper towel on the perforated PVC sheet and then gently place the vacuum bag. I can then run the vacuum suction, but as always, I leave leakage, time to adjust a few pieces of fabric rebels ... Again and again pumpdown deep!! Now, here it is, twelve hours after, you get a nice sandwich very light and very strong!

To follow : the molding of the fuselage, whose technique is a little different due to the complexity of the shape.

Last edited by pirulin; Jan 04, 2013 at 07:26 PM. Reason: comments added under pics
Jan 04, 2013, 07:17 PM
Registered User
pirulin's Avatar
An some more pics....
Weight is good but i calculated that we can theoretically arrive close to 23 gr with this layup
My firs t test was 42 gr, with some mods of process i progress.Nobody is perfect next wing will be better.
Jan 05, 2013, 04:59 AM
Registered User
pirulin's Avatar
Some pictures of the interior of the wing before gluing the bottom and the top surfaces. Why the sandwich technique is useful, because it allows to pass easily guides linkages. For ailerons, I use medical stainless steel capillary tube 0.6 X 0.8 is lighter than plastic guide! ... The control steel rod 0.55mm slides very well. Inox guide (near 20cm + steel wire) = 0.7 gr versus plastic guide + steel wire = 1,4 gr.
The wing spar is made with a carbon socket around a tiny depron spar .
To adjust perfectly the thickness of the spar , I measured the thickness of the interior with plasticine, to know the size of the cutting depron spar and allow closure of the mold with bottom and top surfaces.

Have a nice day !
Jan 12, 2013, 10:24 AM
Registered User
pirulin's Avatar
Molding inside the fuselage on the honeycomb to complete the sandwich:

Like for wing this third step requires the same preparation, but with a difference, this form is complex to laminate. Therefore even wee use cunning. Industrials could make a rigid counter mold or membrane according to the inner dimension of the honeycomb. But we're amateurs, so that's how I do it.

First, once the honeycomb glued inside the fuselage, I sand the honeycomb edges to zero and in any places where there are connections karman. The thickness of the honeycomb is also matched to the drift so that the two half-shells close properly when glueing. Then I put a cotton cloth on the honeycomb and plate well everywhere, I took a marker to delineate the contours of the tissue at the edges of each mold half, it allows me to get a template (once tissue laid flat on a board) that represents the shape of the inner surface of the honeycomb, he remains to trace and reproduce this form on a cardstock.

To laminate the last layer of fabric 25 g/m2 flat on the honeycomb, I took a sheet of wax calibrated 0.5 mm thick with one side adhesive. I cut it as cardboard template. Then I apply non-adhesive side on the honeycomb. Once well pressed, I take off the protective film from the adhesive side and laying strips of PVC sheet 0.3 mm which I cut with scissors and I fit by juxtaposing them to each other to cover the entire surface of the wax sheet. Eventually I apply a few pieces of tape to secure some bands with one another, but not too much, because it requires that all remains flexible to adapt well to all forms. The other side of the sheet of wax is completely covered adhesive PVC film, it is this set which will allow the tissue to be perfectly flat and well plated on honeycomb. Counter mold is perforated with a hot needle all about 5mm small holes to allow drainage of excess resin.

Like the wing, I apply perforated film to drain excess resin, on the wax counter mold side tissue course. Then I prepare the mold by placing the vacuum bag seals etc .... same steps as the stratification of the third layer of the wing. The fabric is laminated on counter wax mold with the least possible to obtain a resin layer is very dry, just wet. I take good care to tender tissue on the leaf wax and remove all wrinkles. Then I put the whole into the mold and plate manually and regularly on honeycomb. The absorbent paper is then placed, other layer of perforated film is placed on the Karman connections and covered with plasticine to press firmly those places. This set is pressed with weight temporarily, time to partial sealing of the vacuum bag. Then the weights are removed one by one and rested on the vacuum bag to the outside as and as the bag is sealed by the seal, there may be a vacuum applied so as to gradually time to adjust the counter mold then the complete vacuum is applied. So, it only remains to wait until the resin hardens for 12 hours at least.

Once the two half shells are quite hard body can finally glued together with resin and microballoon

To be continued.
Jan 12, 2013, 02:02 PM
Registered User
pirulin's Avatar
Its almost finished....

Now the fuselage is glued with resin mixed with micro balloons, bulkheads are manufactured with honeycomb as described before.
I made a bubble with PVC 0.2 mm thickness normally used to make wire bound notebooks: easy to find and cheap. The male mold is polished aluminum. Everything is sucked after a stint in the oven at 165 į c. This gives a bubble of 0.8 gram.
The bubble was temporarily installed on the fuselage and held with the equipment that I made for it (visible in the photographs).
The framework of the bubble is constituted of a wire wound around the edges first with Kevlar and covered with carbon yarn wound around also, the whole is sanded. The bubble is removable therefore interchangeable, four pins allow to fix it, a pin is removable and accessible from the inside of the fuselage.

Control of the elevator is a carbon rod 0.8 mm with 1.5mm ball joint normally used for micro helicopter walkera 4G3 and 4G6.
The elevator control is hidden in the tail of the fuselage and is therefore canít be seen when the rudder is installed.
Three bulkhead reinforcement were glued in the fuselage. For the gear box and motor the bulkhead is sandwich honeycomb + carbon and two others with kelvar + honeycomb.

The battery pack and motor + ESC:

I wanted a removable unit, so it has been realized. There are three fixing points.
The first attachment point enters in a hole behind the bulkhead carrying the motor, the two other are located on the second bulkhead at the leading edge of the wing. A locking device allows both to remove all but also helps retain the battery fastened in its box.
The ESC is mounted on two carbon rods of 1mm with two rubber sleeves to hold.
The gearbox is made in carbon honeycomb sandwich and a tube made with carbon sock allows the mounting of bearings 4 X 1.5 X 2 mm.
The rate of reduction is 5/1.

The holder for Propeller is machined with Dural 2024 alloy.

To be continued with a self-critical assessment of the construction and possibles improvements.
Jan 27, 2013, 03:29 AM
Registered User
pirulin's Avatar
Self-critical assessment of the construction and possibles improvements.

To conclude :

I think this building technique can be really used and suitable for sizes 30 cm and above, concerning the weight, it can still be do better.
The main advantage is that your model will be really less vulnerable to shocks .

On my the first molding, my technique was not well developed, not enough tissue preparation and process = panic = bad result.
I used most of my fabrics by dragging them at 45 degrees so it was too strong and the resulting weight too high.
Another error to be avoided: the amount of resin for bonding the two half shells with the resin charged micro balloons must be minimized.
I think that turning the mold upside down during the resin curing, will permit pump better the excess of resin by gravity and capillary action, so there can still gain weight.
For couples I also,I used the micro ball to the edges, it is not necessary and would save weight. Surfaces indicated for the weight estimate count the top and the upper surfaces for the wing and tail and also the left and right side of the fuselage and rudder.

The first test flights were very good, but because it were too sensitive with the ailerons, the p47 dropped heavily several times without externals damage, just a pawn of the battery pack boken inside: just a gluing again, plus a propeller broken and the shaft of gear twisted.

I now have a project of 15" electric ducted fan jet for which I'll use this technique.

I hope you enjoyed reading this topic and it make you would want to try this kind of building for future projects.

Added on 2013 .02.04 :
I prefer giving you detailed weights because of my errors in the final achievement and reinforcements choosen.

Weigths of the frameís sandwich skins without before assembly (not glued half shells and accessories):

Wing 31.5gr
Fuselage and tail: 19.8 gr
Horizontal elevator : 6.3 gr

Frame Elements finished :

Wing (included former, spar, horns, ball joints, steel linkages and servo mount without servo): 44.3 gr
Fuselage, including bulkhead, rods linkage, ball joints horns, various reinforcements and battery mount (without geared motor and electronics and servos) : 39.9 gr
Elevator with hinge and clevis: 6.6 gr

So a total weigth frame
: 90.8 gr.

I think we can do even more lighter, we can save at least 10 grams!...
As many of us, I always tend to put too much resin with micro balloons when I glued the half-shells (I like to do sturdy ... but at this scale each gram is really important), except for motor bulkhead, the formers are too heavy and strong. As regards the wing, I make same error as the fuselage, I put too much resin and micro balloons. Because the skins are really strong, the wing spar isnít indispensable, on my next models Iíll replace it by studs which ensure the bottom and top surfaces linkage.

Per dm≤ this layup (Kevlar 36gr/m≤ + glass 25gr/m≤+honeycomb 1.5mm+glass 25gr/m≤ gives :

Weight calculated to obtain an estimate before implementation: 2, 2 gr / dm≤
Weight achieved with flat test specimen: 3,03 gr /dm≤

Weight below are "out of the mold (per dm≤)" the halfs parts achieved depending on the complexity of the shape:
For the wing: ~ 3.15 gr/dm≤
For the fuselage: ~3.98 gr/dm≤
For the elevator: ~ 2.86 gr /dm≤

Reading that, you can deduct that the more the shape is complex the harder is to obtain a light frame weigthÖ

Sorry for my english, but I hope it was enough to understand

Have good flights and fun,
Last edited by pirulin; Feb 04, 2013 at 01:29 PM. Reason: english terms added and mistakes corrected, weights added
Jan 27, 2013, 09:55 AM
Registered User
Schoolya's Avatar
Great work

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Kevlar vs glass & cf veil combo FLAME747 Composites Fabrication 5 Mar 01, 2012 01:50 PM
Discussion Glass/Kevlar/CF D-Tube schrederman Composites Fabrication 9 Feb 27, 2012 08:33 PM
Discussion IS cutting carbon/glass/kevlar fabric hazardous? DKNguyen Composites Fabrication 9 Oct 17, 2010 11:20 PM
Discussion Kevlar, Carbon, Composite, Glass with XPS memoryjb Xtreme Power Systems 1 Jul 11, 2007 10:01 AM
Sold Composite Material and Supplies (Carbon, Kevlar and Glass!!) jrbackus Aircraft - General - Miscellaneous (FS/W) 3 Mar 02, 2007 09:51 AM