The DBF 2019 team of FH Joanneum (University of Applied Sciences, Graz, Austria) has won 3. price in this years AIAA Design Build Fly competition in Tucson AZ, USA. Congratulations to each and every member of the team!
Compliments also go to the winning team of Ljubljana University (from our neighbouring country Slovenia) and 2. place team of Georgia Technical University (Atlanta, Georgia, US)
Out of the 104 teams admitted to the competition, 77 teams from US universities and 21 international teams made it into the fly-off. Ranking third among famous name universities like Embry Riddle Aeronautical (6.), Massachusetts Institute of Technology (7.) or Stanford (59.) makes this a remarkable achievement – well done!
What does this have to do with our Sling 4 build? Cobuilder son Paul and his girlfriend Miriam were part of the Austrian team 🙂 .
Being very busy at work, but also because of the Proseal desaster upon our first attempt to close the tank, I have dreaded redoing this step for quite a while. Since then I have worked a couple hours cleaning the remains of the Proseal mess, which brings this little setback to a total of about 45 hrs of wasted build time.
Son Paul has nudged me to go ahead and make some progress again. This time we made a detailed checklist to be sure, that everything was carefully prepared.
I have failed to adjust the travel stops of the fuel sender float arm during our first attempt to close the tank – for the simple reason, that it is not mentioned in the build manual.
It is only described in the maintenance manual (section “REPLACING FUEL LEVEL SENDER ASSEMBLY SENSOR”) – well hidden in the text, where it says:
“The float arm stops (6) on the replacement sensor must be bent to be identical in position to that on the removed sensor“
Of course, in order to be able to duplicate the arm stop position, the arm stops must be bent correctly in the first place. But this is not mentioned in the build manual. Thanks for the heads up, Pascal!
Initial adjustment of the arm stops is easy before closing the tank. It is very difficult (or next to impossible) once the tank is closed – maybe doable with an inspection camera. So this time, after a few iterations I have adjusted the stops thus, that arm movement stops about 4 mm before the float ball touches the skin.
WNR Fuel Sender Float Arm Stops Top Position
WNR Fuel Sender Float Arm Stops Bottom Position
I have also filed a shallow flat notch underneath the locking screw, which prevents rotation of the float arm and gives a reference in case the sender needs to be replaced.
Failing to adjust the stops will let the float arm bang against the tank skin, especially in turbulent air. The resulting vibrations may cause wear inside the sender and early failure.
Some more thoughts about the fuel sender:
I have been thinking about installing a second fuel sender. Because of the wing dihedral, the fuel sender will show a full tank reading from 84 down to 60 l.
I would have prefered to get a sensor reading covering the whole range and it would be fairly cheap and easy to achieve with an additional sensor on the outbord rib of each tank. But these sensors would not be serviceable any more once the tank is riveted to the wing, so I gave up on this idea.
Back to closing the fuel tank:
On Saturday, besides adjusting the float arm stops, we also did all the preparation work once again (Aceton wiping, sotchbrighting, tape masking the mating surfaces once again, Aceton cleaning the rivets …).
Maybe one thing to mention is the Saran wrap, that we used to cover the main spar before clecoing the tank to the wing. The idea is to avoid any Proseal squeeze out glueing the tank to the mainspar. Even just a little squeeze out can make removing the tank again for the leak test difficult once the Proseal has cured.
I will report back here, wether the Saran wrap idea worked.
Edit 07.05.2019 The Saran wrap idea worked well. I had some squeeze out along the main spar and the Saran wrap did avoid glueing the tank to the mainspar. Removing the tank and pealing off the Saran wrap was easy.
After the prep work we could have pushed forward, but we decided to not work into the night and continue on Sunday to actually close the tank (lesson learned from our first attempt!)
This and the careful preparation made closing the tank on Sunday a fairly relaxed endeavour. With the careful preparation we actually needed quite a bit less Proseal this time. Almost exactly one 3 1/2 oz. cartrige did the job this time (with just a little leftover).
Doing it for the second time now, I definitely can recommend using the vertical orientation of the wing as described in my previous log.
With the wing in a vertical orientation it is much easier to cleco and then rivet the tank – compared to working overhead from underneath. Also putting the tank in position was much easier. The tank slid into position under its own weight at first attempt without any wiggling or pushing.
I was afraid, that squeezed out Proseal could block the wire mesh of the fuel pickup line, the return line or the tank vent line. Paul had the idea, to blow out these three lines with compressed air before putting the tank in place, which should clear the lines and blow the Proseal into the fillets. They all seem to be unobstructed.
Sorry for the blurry photos taken with my inspection camera. It is fairly difficult to keep the inspection camera still, get it focused and set the illumination correctly, all at the same time.
For a reference I have given the dimensions of the little blowby hole (2mm). Take notice, how the compressed air blew away the Proseal oozing out from the fillet. It seems like the idea actually worked as intended.
The tank will now remain in place for some time till it gets warmer to allow the Proseal to fully cure. I will then remove it for a leak test and an upcoming inspection of the wing by the Austrian authorities (Austro Control), before finally riveting it to the wing.
Feeling much better now and hopefully the tank will not leak!
Son Paul and I wanted to close the right wing fuel tank. In order to do this, first the vent line tube had to bent and the vent line fitting had to be mounted.
Thinking I had seen a tube bending tool with multiple bending radii in one of our local hardware stores, I had asked Paul to get one and some 6mm Aluminum tube as a practice piece for bending the tank vent line.
Getting the practice tube (4€ for 2m) was easy, but the bending pliers were all the same Chinese make (just in different colours) with a single bending radius much to big for our purpose. Since we wanted to continue with our work and make some progress, I fabricated a bending form by filing a notch in a piece of wood
Clamping the tube with the bending form in my vice and progressively adjusting its position worked best; bending it in one stroke did not work so good. After three practice pieces, I was able to correctly bend the flared piece of tube, that came with the kit.
It has been bugging me for some time, that – when closing the tank – the proseal may squeeze out and possibly block the inlet tube and the vent line, which would be next to impossible to repair, especially the vent line.
In search of a solution Paul and I finally agreed to cover the inlet and vent with some protective plastic with a pull-out-string attached to it.
Don’t do that – at least not on the vent line!
Paul had already primed the Z brackets and riveted the nut plates to the Z brackets, that let you bolt the tank to the main spar. As advised by Sling 4 builder Craig M. (Thanks Craig!) we took care, that the AN3 bolts could easily be screwed into the Z-brackets – some of them needed to be enlarged a little. With the vent line in place we clecoed the rear channel and all the Z brackets to the tank for a final test fit.
For mounting the fuel tank we have decided to turn the wing vertical with the leading edge on top with the following ideas in mind:
any Proseal squeeze out will sag back into the fillet
access through the inspection panels to install the AN3 bolts will be much easier
the tank will fall into position under its own weight
We just carried the wing outside the garage, put the workbenches aside and placed the wing vertically on my trusted wooden building blocks (two of them screwed together on each side).
The vertical position actually worked very well, but a word of caution here:
I would not use this method without all wing skins in place! Since the sides of the jigs are not perpendicular to each other, rotating the wing will cause the wingtip jig to not fully rest on its support. The rear top corner of the wingtip jig will be up in the air.
If the wing skins are in place this will not matter. Since the wing is so torsionally stiff, it will not change twist at all under its own weight any more. If you take a close look at above photo, you will note some wooden wedges we have put in place nonetheless to support this corner.
But if not all wing skins were in place I am pretty sure, that the wing will (at least to some degree) change the wing twist under its own weight – definitely not a thing you want!
The final testfit went quite well, just the outboard most Z bracket needed to be pushed a little into place using hole scribe to easily get the AN3 bolts in.
Up to here everything worked out quite well and we considered ourselves ready to mix the next proseal cartridge. After reworking a couple spots inside the tank just for peace of mind, we riveted closed the tooling holes of the end ribs (ribs 101 & 105) using M5 washers. The washers are needed, because the grip length of the 4,8x10mm rivets supplied is too long for the thin rib. This is mentioned somewhere in a text page of a newer version manual, but not shown in the drawing.
We prosealed the vent line fitting in place, but we did not attach the ventline yet as riveting the Z bracket and rear tank plate to rib 105 would be difficult with the vent line in place.
With that the fitting done, we started closing the tank. I do not want to get into any further details here other than saying, that we decided to unrivet the rear tank plate again next day. Fortunately (with the low temperature) the Proseal had not cured yet and removing the rear tank plate again was fairly easy.
But cleaning up the Proseal mess was not fun at all!
Edit 19th February:
Skiing in Bad Gastein with wife Elisabeth, daughter Vera and son Paul with four absolutely cloudless days full of sunshine and snowcovered mountains all around definitely make for a nice break from Proseal.
Makes me looking forward to be able to fly my own Sling4 over snowcovered Alps during winter on a day like this some time.
It’s been a while since my last post and building has slowed down but I have been able to make some progress.
Firstly I have continued riveting the right wing skins. which now leaves the wing tank and the wing tip missing on the right wing.
As I have indicated in my last post (“So let the Proseal fun begin!“) building the wing tanks is a somewhat dreaded task. Actually I am not using Proseal™ but Flamemaster™, another CS3204 B2 compatible sealant recommended by TAF.
Especially with the last step (closing off the tank – which I have not done yet), I have to admit, that I am still a little anxious about. Once done, there is no real good way of sealing leaking tank – at least not an easy one.
Some good tips from the Van’s community about, what should be taken care of to properly seal the tank during the build, have been around for quite a while. I am adding three pictures here, which summarize these pretty good.
The first picture shows, where one should take a close look, when sealing the tank. Adhering to these recommendations normally provide a dual or triple redundancy and avoid single point failures.
The second and third picture give recommendations about the shape and size of fillets.
Instead of the catheter syringe method, that TAF suggests, I have decided already at the beginning of my build to use a Semco pneumatic sealant gun for applying the sealant.
Used sealant guns are readily available on Ebay in the US starting at 30$ (add customs, import sales tax and shipment … and you can get it for around 80€ in the EU 😦 ). I received mine in pretty good condition, it just needed some cleaning. From my experience so far, I definitely would not want to build a fuel tank without it.
As opposed to information (given somewhere in the FAQ) at Aircraft Spruce, the Flamemaster™ (sometimes also called CHEMSEAL ) CS3204 B2 3 1/2 oz Semikit cartridges (available at Aircraft Spruce), that I am using, DO fit and can be used with the Semco sealant gun.
The sealant gun, together with using Semikit cartridges makes mixing and applying the sealant much less of a mess. In fact I have managed not to have a single sealant smudge on my work clothes so far ;-).
I have modified the nozzle of the Semikit cartridges by bending (softened it with a heat gun) and cutting of the tip, so that it
– fits inside the stringers and
– easily fits over the 3mm rivet heads.
Doing so made it easy to encapsulate the shopheads with the sealant gun. This would be time- and labour intensive when done by hand. Just place the gun over the rivet head, shoot , allow for a little squeeze out under the tip of the nozzle and back off – no manual postprocessing needed.
The curved nozzle makes handling the rivet gun in the limited space much easier, once the tank is in the jig. It even works when encapulating the rivets inside the stringers.
To spread out a sealant bead, I just used this.
One important recommandation can be found in the manual:
CLEAN – LIGHTLY ABRADE – CLEAN AGAIN
which to my understantding should be done within 12 hours of applying the sealant.
In the technical datasheets I have found MEK (Methylethylketon) and Toluene as recommended cleaning agents for the CS3204 B2 type sealants and TAF is using Benzine for cleaning – none of which looked appealing to me for indoor use.
Sling 2 builder friend Pascal L. had contacted the manufacturer of Flamemaster and they ok’d using Acetone . Acetone to me seems to be the least hazardous, or in the words of Pascal: “Hey, our girls have used it all the time as nail polish remover, so it can’t be that bad”.
Pascal, who is just slightly ahead of me with building the tanks, was kind enough to share his experience he made during experimenting and building his tanks on what worked well and what did not work so good for him. This also helped and propably saved me a couple of tank sealant cartriges – thanks Pascal!
Proper preparation is essential. The 2 hours workable time of the CS3204 B2 don’t allow much for getting distracted. Here’s what I have taken care of:
Make sure that a rivet fits in each and every single rivet hole during testfit
Make sure that each piece of hardware fits into its corresponding hole
Clean-abrade- clean mating surfaces with acetone and scotchbrite
Soak rivets in acetone, clean and dry with paper towel;
Same with a sufficient number of clecoes
Clean rivnuts with acetone and dry with paper towel
Same for any hardware, stiffener rings, etc.
Mask off with masking tape around any area that will be sealed but leave 6mm unmasked, that need a fillet.
Have enough (~200) nitrile gloves (non powdered) at hand. Make sure they don’t dissolve in Acetone. If it has sealant on it, take a new one!
Have a couple rolls of paper towl at hand
And propably the most important advice:
Have someone to help building the tanks
On my first batch of proseal I had installed the right tan stringer, cap flange, drain flange and sender flange plate on my own. I had son Paul and his girl friend help (Thanks!) when working with the second batch of proseal to the attach the ribs to the tank skin.
It does make such a difference and takes out a lot of stress, if you have someone helping to clean the squeeze out, remove masking tape, apply missing masking tape, continue shooting rivets, while encapsulating the rivet heads myself, …… and so on.
Here are a couple photos of the process. Some areas still will need some finishing or rework, which will be done when closing the tank with the back plate WG-CHL-009-R-C-
Having a Jar with Acetone at hand for used clecos makes later cleaning of the clecos easier.
Instead of countersinking the cap flange (the TAF supplied 4mm sealed rivets have insufficient griplength without the countersink) I used 4×12,5mm sealed rivets on the cap flange.
I have also used sealed 3,2x 12 mmV2A stainless steel rivets for the fuel drain flange.
I am using M4 sealed rivnuts for the fuel sender flange plate (instead of the TAF supplied unsealed) and will attach the fuel sender mounting plate with M4x12 stainless steel hexhead screws.
While letting the sealant cure, I temporarily clecoed the backplate in place.
After the sealant had cured (on my third proseal batch) I installed the rigth wing fittings (except for ventline – see below), lid flange of the locking fuel cap and the left wing stringer and flanges.
The manual wants you to mount the fittings for fuel outlet, fuel return inlet (attached to rib 101) and fuel vent(attached to rib 105) before rivetting and sealing the tank ribs to the skin. If done this way, there would be no way to make a good fillet behind the wire mesh inlet and the fittings. Therefore I prefered to do it after the ribs were in place and the sealant had cured.
The anti-rotation bracket will need some more sealing . Also visible in photos above and below are the M4 sealed rivnuts for the fuel sender flange plate, that I am using.
To align the lid flange with the tank skin I fabricated this clamping device (Thanks Pascal for the idea!). Without clamping, the lid flange will slightly sag and not align with skin.
I have tried to rivet closed the tooling holes in ribs 101(upper right and bottom center in photo below) and 105 (near return inlet in above photo) with the 4,8mm sealed rivets called out in the manual. The grip length of those supplied with the kit is too long for the materialthickness and just will rotate freely. I will either have to get shorter ones here or add an M5 washer.
Last fall I had aero club fellow member and aircraft mechanic Patrick (working for the Red Bulls Flying Bulls ) visiting. He was inspecting my build progress and we were also talking about the tanks being next in my build – specifically about the cork seals (I will address this in one of my next blog entries).
While talking, he took a look at the 90° bulkhead fittings. The AN6 bulkhead fittings for the fuel lines looked good to him, but he advised me not to use the AN4 bulkhead fittings for the vent lines.
Patrick then told me of an in flight fire and the complete loss of aircraft that was caused by this very type of bulkhead fitting. He then explained to me, what had happened and why. Thanks for the heads up, Patrick!
Have a look at the correct AN833 fitting (left in photo below) and the fitting in question (right).
The incorrect fitting has a very narrow collar. When installed with the required washer (shoulder is too narrow to install without), the fitting can move freely in the its hole.
In the accident mentioned above this fitting vibrated loose and fuel ran down on a hot exhaust. Admittedly this was in a high pressure fuel line, high vibration installation – which is not the case here. With the AN833 the fitting cannot move, because the collar is a tight fit to the hole, so this is cheap insurance.
Since I did not yet have the AN833 in hand when working on my third proseal batch, I will have to install them on the next one. So far working with proseal on the fuel tanks was not that bad but only the leak test will ultimately tell the truth. Next will be installing the fuel tank back plate .
In a meeting of IECA (Igo Etrich Club Austria, the Austrian counterpart of EAA) I sought approval from my Austro Control technical examiner (the Austrian authority) to put the wing skins on. He ok’ed this as long as it was possible to inspect the interior with a mirror / borescope.
With all the prep work done, putting on the right wing skin was straight forward and easy. The only prep work left to do was to apply Ardrox AV30 to the mating surfaces with a foam paint roller, which I did immediatly before clecoing the skins.
One thing I did different compared to the testfit was to start with first clecoing the overlap of the top skins to rib 8 where I had some tension during test fit. It was much easier and no noticeable tension anywhere.
There is a line of three unused rivet holes behind rib 4. Mike at TAF confirmed, that nothing is attached there and that they should just be riveted closed.
Since I knew that I would not have enough hands to turn the wing upside down in order to rivet the bottom skins, I have raised the wing jig off the table a couple weeks ago by about 25 cm.
Using three boxes strategically placed underneath the wing structure to lift the bottom skin off the table close to the underside of the ribs made it easy to cleco the bottom skins in place myself.
I had to rivet the whole bottom skin in a bent forward and often twisted, overhead position, which – considering the number of rivets – was somewhat inconvenient. Stretching the work over a period of three days and frequently changing position this was ok.
It was quite a satisfying feeling to see the skins on. The fit seems to be quite good – no noticeable bulging or waviness.
As mentioned above, my technical examiner wants to be able to inspect the interior of the wing. This would not be possible at the wing step ribs once the tank is on. We therefore agreed on having the next inspection once the tank is finished but not yet riveted to the wing. So let the Proseal fun begin!
After fabricating the conduit support brackets for the wing ribs, I finally was able to rivet the wing ribs to the main and rear spar.
I had a little problem with the flange of wing rib 16 that attaches to the rear spar being bent slightly incorrect. Due to that the rib was 1,5mm too short pulling the main and the rear spar together. This put a lot of tension on the rivets and resulted in some waviness of ribs 15 to 12. This effect would not show with clecos, because the clecos were just pulled out of the rivet holes a little. Putting a 1,5mm shim between rib 16 and the rear spar solved this.
Peter Calley has reminded me not to forget to put the flap torque tube in place (thanks!). It cannot be inserted once rib 6 is in place unless you unrivet the lever arms.
One thing I also have done is, to replace the old style 3mm Nylon aileron pushrod guide (that was just riveted to rib 6) with the new style Vesconite bush, that TAF supplies now – much more durable IMHO. Since my rib 6 did not have the holes for the new bracket predrilled, an installation drawing Armin from TAF sent me was a big help .Before drilling the rib I checked the bracket position by temporarily putting the long aileron push rod in place connected on one end to the aileron bellcrank and centered in the rib 12 hole on the other. For doing this, I had to make hole in the jig. Some grease helps with threading the long push rod. All the dimensions given were spot on, so I match drilled the holes and riveted the bracket in place.
Installing a short stretch of conduit for a test, it quickly became obvious that it flopped around. There is no good way to support the electrical wiring (or conduits in my case) between the ribs. Builders have come up with various solutions.
I have decided to make support brackets attached to the short vertical stiffener channels on the main spar using existing rivet lines similar to the ones Sling 2 builder Pascal L. has documented here (thanks Pascal).
The following photos show some of these support brackets and the routing of the installed conduits.
I have installed and torqued the eyebolt bearing now for the torque tube and the aileron bellcrank already now . The manual does not suggest that, but I think it is much easier doing it beforehand on the workbench than later through the inspection panels.
Another thing I should have done at that point is to chamfer the upper and lower wing stringers, so they sit nice and flush on the skin at rib 9. Obviously I have not and I later had to later remove the inboard bottom skin once again, after it had already been clecoed in place for riveting 😦
Scotchbriting the skins with Prekote and priming with Adler 2k epoxy primer also took some time. Space is getting somewhat tight.
As mentioned in my last blog entry, I wanted to have a look at how the fuel tank fits, before taking the wing skins off again to assemble the right wing structure.
To assemble the tanks, a jig is required. The manual has the templates for the jig, which I printed out using the poster option with alignment cross hairs of Adobe Reader to get dimensionally correct templates. The dimensions given in the plans allow checking this.
I am posting this tip here, since some fellow builders mentioned having a problem printing to scale.
For padding the jig to prevent scratching the skin I used a narrow stripe of carpet glued to the jig with contact glue.
Clecoing the tank together was quick, everything lined up nicely. I just had to ream out the 3,2mm rivet holes – either the holes were just slightly too small or the batch of rivets too large.
My wife helped me with my first trial to testfit the tank. But whatever wiggling, pushing or pulling we tried, I just could not push the tank fully towards the main spar and had a gap at the tank / nose skin overlap and the top main spar rivet line did not align either. So off it came again.
After some head scratching and measuring I found out, that the outboard most tank mount bracket that let’s you bolt the tank to the main spar, was sitting high on a rib flange. This made it impossible to push the tank fully onto the main spar. This was not obvious, as this area is closed off, when the tank is mounted. Even a miniature borescope would not get you in there. Turning the mount bracket 180° fixed that issue.
As expected, I had some misalignment at the bottom tank / wing skin overlap – nothing serious – a little elongating of a couple rivet holes fixed it.
One hole in one of the tank mount flanges near the wing root needed some filing too, but that’s ok for me, since the AN3 bolt used here will need a washer under the bolt head anyway.