Supra spar questions, it was broken... :)

[Berrie]

I’ve just started to build my Supra. I’m now making a materialorderinglist for EMC-Vega. This is a good supplier of materials here in Europe. I arrived at the point what to do with the spar. I’ve two questions.

1) EMC has a hughe amount of sleeves (schlauchen), so I cant make a choise wich one to use for wrapping the spar. The sleeve has to replace the two bias 54gramsquaremtr wraps arround the centerspar and one layer at the tippanels.

http://www.emc-vega.de/pdf/mat07.pdf

can anyone give me an advice? I’ve used SC 3200 before, but that seems very have to me for this glider… Iwould like to use the lightest sleeves possible. So centerspar en tipspars would have a different sleeve. When there's nothing suitable I will use the by Mark Drela recommended bias wrapping.

2) I also want to use HTS rovings instead of the recommended ACP “ready to glue”spars.
Is it possible to take the local carbon spararea an divide it by the rovingarea to determine the number of rovings? (24K, 1610tex)

Thanks for helping!

[markdrela]

1) EMC has a hughe amount of sleeves (schlauchen), so I cant make a choise wich one to use for wrapping the spar. The sleeve has to replace the two bias 54gramsquaremtr wraps arround the centerspar and one layer at the tippanels.
can anyone give me an advice? I’ve used SC 3200 before, but that seems very have to me for this glider… Iwould like to use the lightest sleeves possible. So centerspar en tipspars would have a different sleeve.
The attached PDF shows the loads on the spar wrap in the center, where the wrap must transfer the load F into the rest of the wing.
Using an F3J 250 lb launch load requirement on the tension-only skin case, we have

F = 125 lb (front and back of spar wrap each carries half the 250 lb total load)
h = 0.77 in (Supra spar height)
sigma*t = F/h = 162 lb/in

Assuming a max tension stress of 200 ksi for carbon (conservative, allows for fracture over the spar corners, etc), I estimate the following safe max loads for carbon rovings, or "tows":
3K tow: 60 lb
1K tow: 20 lb
So the spar wrap material will need less than 3 tows/inch for the 3K tow, and less than 9 tows/inch for 1K tow. The SC 3200 is made from 3K tow, and surely has more than 3 tows/inch, so it's serious overkill. But the SC 1070, made from 1K tow, seems about right. For insurance, maybe add a second layer over the center 1/3'rd of the center spar.

The required spar wrap strength decreases roughly linearly from center to tip, so the
SC 1070 is far stronger and heavier than needed for the tip panels. The lightest solution here will to to use 1.5 oz (50 g/m^2) glass. Don't forget extra wraps over the joiners.


2) I also want to use HTS rovings instead of the recommended ACP “ready to glue”spars. Is it possible to take the local carbon spararea an divide it by the rovingarea to determine the number of rovings? (24K, 1610tex) Yes, this will work OK. I would use the highest-modulus roving that you can get. With the 3/4" x 0.084" center caps, a high-modulus tow will easily make the spar stiff enough for the strongest F3J launch.

[Christian Baron]

Berri,
I'm living close to Recklinghausen where EMC-Vega is located. I've also build a Supra wing, but extended to 4,5m!
My spar is laminated from 12K rovings. The number and the reduction in size I have calculated with my small excel-sheet you can find in the download area on my homepage.
I've used the SC1062 braid from EMC-Vega for the center spar web. In the middle two layers for the first 15cm from the center. The braid was laminated on a PE-film and then wet adopted to the Spar and vacuumed.

The outer wings I've build with a 49g/m² glass wrap like Mark has recommended.

Christian

[Christian Baron]

Here two more pictures of my wing.
Christian

[Berrie]

@ Mr. Drela.

Thanks for your answer.

Reaction 1)

About the SC 1070. It has his tows at 45° at a diameter of 45mm. The Supra spar had a diameter from about 20mm.
In the PDF file you showed an 45° example. The ideal situation. It is by the way a very clear explenation!
But after a night with (again) a heavy thunderstorm in the Dutch sky something passes trough my mind.

For my own understanding:
Reactionload F’= 0.7*Fapplied load->comes from 0.5*F applied load*sqrth2.

That makes:
If the wraporentation is not 45° but, for example, 30° (to horizontal) then the reactionload F’=0.5*F apllied load*2 (triangel 1-2-sqrth3) That gives a 30% higher reaction load F?! This might give a problem when I use the SC1070. Because the tow orientation. Or am I mistaken mightily?! Correct me if I'm wrong!

About not forgetting the extra wraps arround the joiners. I’m used to wrap them with kevlar tow. The Genieway. I guess both ways will reach the same? Or has the extra glaswrap an other reason?

The tipspars shall be wrapped with G54 since it has 24*24tows instead of 24*19 for G49

Reaction 2)

Great. I’m happy that is possible to use the HT rovings. Then, the main purpose for HM rovings is stiffnes? I’m not flying F3J contests. How can I say it, about me and thermalflying. Let’s say it this way: “I’m an lover, not a fighter”

So, a quick look to HM and HT roving. The HM is only 14% stronger, but 82% stiffer and unfortunaly less ductile.

Is it possible to make the spar more gorillaproof by thicken the upperspar with, lets say, 15%?

However, I will inform EMC about the dayprice from HM roving.

(my first idee was to buy 2kilo HT roving since it is over many years the cheapest solution. I can use it for many things. Local fuse reinforcement, spars and joiners.)

@ Christian.

Thanks for reacting!

Ein 4.5mtr Supra. Was fur ein tolles idee! Ich glaub du bist der erste Deutscher mit ein eigenbau Supra?! Auf RC-network.de habe ich noch keine gesehen..

I was a few months ago near to Recklinghausen. Its somewere between my house and the Wasserkuppe.

I will use your excelsheet for “control”.

Thanks for suggesting the SC1062. It has the tows at 45° with a diameter of 20mm. It must fit the original Supra spar.

About the pictures. How will you epoxy the enormous centrepanel? I’m already worrying about epoxying the big original centrepanel. The biggest centrepanel I’ve ever done had a span of 1.3mtr and it was almost an hour of hard work! If you have a suggestion for relax epoxying suchs a big panel………..

Off topic, butt I’m curious about flying pictures of your Supra…

P.S. I’ve started an buildtreath at an Dutch forum:

http://www.modelbouwforum.nl/forums/showthread.php?t=43432

[Christian Baron]

My extended Supra is not finished just now but will be in some weeks. The center panel is ready and covered with 93g/m² carbon biaxial. I've used the technique Mark Drela has provided in his articels and threads. The mylar film I've bought from R&G. Just working on covering the outer panels.
It is not an original Supra what I'm building. For an original Supra I have cores cnc cutted waiting for finishing my large CB62 first.
Only the wing span is increased not the width.
Here a picture of my drawing.
If you would like to discus with me a little bit more sent me an e-mail and I will provide you my phone number.

Christian

[markdrela]

For my own understanding:
Reactionload F’= 0.7*Fapplied load->comes from 0.5*F applied load*sqrth2.
That makes:
If the wraporentation is not 45° but, for example, 30° (to horizontal) then the reactionload F’=0.5*F apllied load*2 (triangel 1-2-sqrth3) That gives a 30% higher reaction load F?! This might give a problem when I use the SC1070. Because the tow orientation. Or am I mistaken mightily?! Yes, the wrap fibers will be more highly loaded if they are at 30 degrees rather than 45. I would just use a second layer over the center 1/3 or 1/2 of the spar. This will also give a more optimum wrap strength variation over the span.


About not forgetting the extra wraps arround the joiners. I’m used to wrap them with kevlar tow. The Genieway. I guess both ways will reach the same? Or has the extra glaswrap an other reason? Compared to most of the spar, there are two additional loads at the joiner:
1) A much larger shear load. For a 250 lb wing load, the sparcap load at the joiner is 1250 lb. The resulting shear load on the 0.62" x 2.5" joiner is 1250 lb * (0.62/2.5) = 310 lb, which is more than the shear in the center of the wing. Circumferential Kevlar wrap does not help here, since it cannot take a shear load.
2) A bursting load from the joiner rod acting as a lever on the joiner box. This bursting load is the same 310 lb as the additional shear load calculated in 1) above. The shear wrap can withstand some of this load, but it's best to add a few Kevlar or carbon wraps like you've done. A few 3K carbon tows is enough. I add these only at the very end of the spar, where the bursting stress is most concentrated.



So, a quick look to HM and HT roving. The HM is only 14% stronger, but 82% stiffer and unfortunaly less ductile. Is it possible to make the spar more gorillaproof by thicken the upperspar with, lets say, 15%? The Supra sparcaps are sized entirely for stiffness. At a 250 lb load, the sparcap stress is only 75 ksi (520MPa), which is less than 50% of the failure load. To break the spar, it would be necessary to bend the wing into a U-shape, with the tips vertical. So strength is not a problem even with the "weaker" HM roving.

[Berrie]

Thanks for the answers! :)

[Berrie]

To tell de truth........

I'm back on glasswrap.

It started with the shocking fact that I bought 10inch to short carbonsleeve.

After I recalculeted the weight it came out that the carbon wrap is about 10-12grams heavier (exlusief epxoxy!!)

So, knowing that, and learning to deal with the sticky biasglaswrap, I decided to use de recommended glas.

In combination with the HM spars I think the spar has an good stiffnis weight ratio. The spar came out at 162gram. That scares me.....
Some photo's:
http://www.modelbouwforum.nl/forums/zweefvliegen/43432-de-bouw-van-een-supra-9.html#post806049

[markStockton]

Yes, this will work OK. I would use the highest-modulus roving that you can get. With the 3/4" x 0.084" center caps, a high-modulus tow will easily make the spar stiff enough for the strongest F3J launch.

How do you calculate the number of tows required for a high-modulus F3J Supra spar? If the rolling black outs ever stop I intend building a super stiff HM Supra for F3J as I believe a lot of line energy that could be used for faster / higher launches is being wasted in bending the wing's.

[jfrickmann]

Mark - I am building a 3m electric inspired by Supra and Aegea. I am planning to use aramid sparwrap - 1.7oz on the center and 1.0oz on the outer panels - any problems with that?

Christian - is that CB62 your new Sunset model - with the Reisenauer 13:1 gearbox and that giant 30" prop? I was looking a little at your homepage a month ago - that's really cool :cool:

Thanks,
Jesper

[Christian Baron]

Jesper,
you are right, this model I will fly this year at the Sunset Cup. Weigth will be 2400g i hope, because it is not finished jet.
I have not build it only for this special contest, I will have a second nose cone with a much more powerful motor and 5sLipo's for fun flying.
Christian

[atjurhs]

Mark - I am building a 3m electric inspired by Supra and Aegea. I am planning to use aramid sparwrap - 1.7oz on the center and 1.0oz on the outer panels - any problems with that?

Jesper, I've used this method on open-bay wings in the past and it's held-up very well.

On the "Supra-Like" spar I'm working on right now I plan to again wrap with kevlar as you see pictured and then 1 layer of FG cloth on the bias. I have no "real data" to back me up, but based on the past, I think this is very safe, and much cheaper than 1.7oz aramid cloth.

[Berrie]

How do you calculate the number of tows required for a high-modulus F3J Supra spar?

I use 80gr.m^2 HM UD carbon. No rovings. During building the firts tipspar I learned that 1 layer resulted in a thikness of 0.09mm. Number of HM UD "sheets" = the recommended 2mm/0.09 = about 22 layers.

My spar is at the center 18*2mm to 18*0.5mm at wingjoiners. The 18mm is just practical because of the dimensions of the woodparts I've used to press the carbon sheets together. More ore less the way Cristian builds is spar.

http://www.modelbouwforum.nl/forums/zweefvliegen/43432-de-bouw-van-een-supra-8.html#post786446

[markStockton]

Thanks Berrie, I have a roll of higher modulus 12k carbon tow and would like to learn the method of calculating the correct sizing of the spars to make a very strong / stiff wing. I love the Supra design, however molded product is just not nearly strong and stiff enough for serious F3J work.

[nuevo]

I suggest making a small sample piece using say 20 pieces of your tow. Use your usual construction methods. Afterwards, measure the cross-sectional area for 20 tows. Knowing that should tell you how many tows to make any cross-sectional area you desire. HTH

[Christian Baron]

Mark,
you can use my excel-sheet you will find at the download site at my homepage. You have to know the number of filaments of your tow (we are call it roving). For example:
12k = 12000 filaments = 800 tex, have a cross section of 0.7mm²
24k = 24000 filaments = 1600 tex, have a cross section of 1.5mm²
36k = 36000 filaments = 2400 tex, have a cross section of 2.1mm²
You will find one of these data on the tow spun I hope.
Christian

[markStockton]

Thanks CB and Nuevo, good idea.

I was also hoping to get a comment from Dr Drela on the wing flex wasting launch tension and thus reducing the efficiency on an F3J type launch ;)

[nuevo]

this post is probably what you were looking for. I used the advanced search on all posts by markdrela and the keywords f3j, flex, and launch.

http://www.rcgroups.com/forums/showpost.php?p=8745069&postcount=124

[Phil Barnes]

Nice search John, however that post is talking about RDS shaft flex. I re-ran your search without the term f3j and found the following:

http://www.rcgroups.com/forums/showthread.php?p=6357117&highlight=flex+launch#post6357117

That one is directly on point. It mentions that his Supra design could gain at most 0.5 meters from the extra lift of a stiffer wing on launch but would lose 4.6 meters from the extra weight of the stiffer spar.

Applying that to Mark Stockton's case though you would need to recognize that Mark Drela's Supra wing is designed to flex without twisting. A wing that twists significantly under a bending load would lose a lot more launch height (or worse) from extra wing flex. Also, the weight penalty for doubling the Supra's spar was based on the weight of CST pre-cured laminates. The weight penalty for doubling the stiffness of Mark Stockton's spar would be based on the weight of his HM carbon tow.

I just went back and reviewed this thread before sending this post. Mark Stockton is building a Supra wing so if he wants to go with a wing that flexes as much as the original, it would be important to accurately reproduce the wing planform and sweep angles, the spar location, and the correct bias fabric layup (to maintain the correct torsional stiffness of the wing). Those are all important when it comes to obtaining a pure wing bending under winch loads without significant wing twist.

[nuevo]

Thanks for clearing up my goof, Phil. I skimmed Mark's posting too quickly.

[markStockton]

Thanks Phil, that was exactly what I was looking for. You also reminded me about the importance of torsional stiffness. This is actually the big problem with the molded Supra's we've had, they weren't stiff enough, thus loosing out a lot on launch. We destroyed two during our national F3J team practice this weekend. Both where complete torsional failure of the center panel and inner portions of the tip panels. The wing's look like wavy potato chips.

My intention with the HM foam wing is to follow Dr Drela's planform, as well as use stiffer materials for torsional strength. I was planning on using 68 gram carbon for the center panel, and carbon / kevlar hybrid on the tip panels.

Finally I must mention, I'm not terribly happy with the way model aircraft manufacturers take no responsibility for under spec'ing structure. One of the Supra's that went down was less than a year old, and was supposed to be the beefed up F3J wing. In mountain biking the 2008 Cannondale model has been recalled due to frames cracking while being raced. The factory is shipping replacement frames to all effected customers at the factories cost. The production Supra is not a cheap model, yet we continue as customers to fund the R&D of the model. I must say I think it is BS.

[Kiesling]

Thanks Phil, that was exactly what I was looking for. You also reminded me about the importance of torsional stiffness. This is actually the big problem with the molded Supra's we've had, they weren't stiff enough, thus loosing out a lot on launch. We destroyed two during our national F3J team practice this weekend. Both where complete torsional failure of the center panel and inner portions of the tip panels. The wing's look like wavy potato chips.

My intention with the HM foam wing is to follow Dr Drela's planform, as well as use stiffer materials for torsional strength. I was planning on using 68 gram carbon for the center panel, and carbon / kevlar hybrid on the tip panels.


Hi Mark,

I'm a little surprised you have a problem with the torsional stiffness of the molded Supra. I have seen the glass ones launched quite hard and did not think they were giving up anything on tow. I had the Kevlar version and it was fine in torsion. With the straight spar design of the Supra, the torsonal loads are lower than on models with sweep. So the torsional stiffness requirements are lower.

The only model I can remember having a torsional failure was with a molded Mantis. The rear shear web failed. It was made of balsa runing span wise without glass reinforcement. It split along the grain during the pull up from the zoom. Wrinkled the skin pretty bad. The friend of mine who built it for me replaced it . . .

I'm wondering if it could have been a flutter induced torsion failure. If you are launching really hard and flutter is initiated, it can be quite violent and a lot of damage can occur. I know a lot of the earlier molded Supras had flutter problems. Most of this was from the surfaces being too heavy and lowering the natural frequency - not necessarily that the surfaces were not stiff enough. In fact I think most of the problems happened with the early carboned skin versions and not the glass ones (my memory is a little fuzzy on this).

Anyway, if you are making a foam cored version, I think using the Kevlar lay-up Mark Drela shows on his plans is adequate. If you wanted a little more torsional stiffness, it shouldn't hurt to use the 68 gram carbon on the center, I think this option is on the plans. For the tips, what weight carbon/kevlar cloth are you planning on using? My bagged Supra wings use 1 layer of 2 oz/sq yard carbon/Kevlar hybrid for the entire wing. It isn't the lightest, but it is pretty stiff torsionally and I haven't had any noticeable flutter problems.

Tom

[markStockton]

Here the photos I took for Vladimir, it is hard to believe all this damaged occurred in the air before the model hit the ground.

An interesting point on boom construction. It seems that Vladimir uses bi directional carbon / kevlar cloth for the boom that is oriented at 0 - 90 degrees. This means 50 percent of the material is serving little or no purpose.

Tom, I think you are right, this might be the result of high speed flutter. However this isn't an early model, it was #173 and thus had the extra carbon in the ailerons as well as the taller aileron linkage, sub ribs etc.

On my foam wing I was planning on using 70gram carbon / kevlar hybrid for the tips.

[Kiesling]

Hi Mark,

Thanks for posting the pictures. These pictures remind me of a failure of a Calypso Cobra with an F3J layup being launched on an F3B winch. The skin on that model was all glass and did not have the flutter resistance needed for the speed generated using an F3B winch. Servos were ejected in the air and it was quite spectacular . . .

I think what happened to this Supra is flutter failure. From the pictures it looks like the layup on the flaps is on the heavy side. It looks like the mode shape for the flutter was the entire center panel - not just the control surfaces. So you could have a super stiff control set up with ribs and all, but it wouldn't matter if the weight/stiffness distribution allowed the flutter frequency of the entire structure to be reached during the tow.

The carbon doubler on the flaps should not be as big as it is. If you look at the Supra plans on the CRRC web site you can see that the doubler on the control surface is less than half of what is shown in your pictures.

At what point in the launch did the failure occur?

[markStockton]

At what point in the launch did the failure occur?

The plane went as I pushed over into the zoom.

I agree on the flutter. When I was taking the photos last night I realised this. Although, the gears in the JR DS-161 being stripped on both ailerons being stripped was a big hint. ;-)

I had been working on a faster launch set-up using the techniques published by Martin Webershock. The result is a very quick launch on our X-pro's. I suspect to fast for a production Supra.

At the end of the day I've decided Kurt should fly X-pro's in Turkey. I'm only the reserve and doubt I will be flying other than the pre-event. Hopefully I'll have my foam core model done by then. It's a pitty you and Phil won't be there. Oh well next time, or maybe the first F3K WC?

[Phil Barnes]

Who says I won't be there? Actually, it looks like I will be towing.

[markStockton]

Who says I won't be there? Actually, it looks like I will be towing.

Cool, me to. Are you going to bring a chuckie. I'm launching a bit higher these days. ;)

[Phil Barnes]

I don't know anyone named Chuckie..... Oh! You meant a HL plane. That would depend on whether there is a convenient box to pack it in. If there is an open F3J contest before the real event and I bring F3J models for that then I'll bring a DLG also.

[markStockton]

If there is an open F3J contest before the real event and I bring F3J models for that then I'll bring a DLG also.

There is an open F3J comp the weekend before the event, much like the Martin Cup in Slovakia, I'm going to be flying in it.

[davidleitch]

Mark

Its interesting to follow through on model failures. You mentioned two failures and one has been discussed here. What happened to the other one and do you happen to know what model number it is?

Like many models supra development seems to continue. For instance I received a newsletter in December which contained inter alia:

3. Supra's ailerons became longer, they are now available across the
full length of the wings tip section.
4. Wings tip of Supra Glass have been redesigned, which reduced their
weight by 7...8 grams (0.25...0.28 oz)--now, they are 18...20 grams
(0.6...0.7 oz) lighter than the wing tips of Supra Carbon Light.
The new design also resulted in the greatly increased strength of the
ailerons.

[dbarker]

The plane went as I pushed over into the zoom...

I had been working on a faster launch set-up using the techniques published by Martin Webershock. The result is a very quick launch on our X-pro's. I suspect to fast for a production Supra.

...



Mark, could you post a link to Martin's published technique for a faster launch set-up.

Thanks,
Don

[jirvin_4505]

Mark sent me a copy - thanks Mark.

I have found it post in a couple of places on rcg with discussion..
http://www.rcgroups.com/forums/showthread.php?p=7940545#post7940545
post #11

and reposted here.....
http://www.rcgroups.com/forums/showthread.php?p=7943534#post7943534

post #6

cheers Jeff

[dbarker]

Mark sent me a copy - thanks Mark.

I have found it post in a couple of places on rcg with discussion..

cheers Jeff

Thanks Jeff

After rereading, I now remember seeing this before with the major point being trimming the plane for neutral flight with launch camber in the wing rather than leaving the elevator in thermal or normal flight setting.

Good to hear this technique resulted in faster launches, bad to hear this resulted in a whole wing flutter issue with a molded Supra.

don

[Berrie]

The nightmare became true. During a heavy highstart the spar of the centre panel broke. Now my eyes are not red anymore from weeping, its time to recover and built a new centrepanel as fast as possible.

But first time to reflect on what actually could be wrong.

The circumstances during the fatel highstart: Full pedal and +/- 4 m/s windspeed. It broke at an heigt of 100 mtr. I was pulling the stabilo (a lot…).

The spar broke at the right site of the centrepanel. At 330mm from the centre.

Some information about the spar:

The sparcaps are built up from layers HM carbon. The glasthread that holds the fibers together is still at the carbon. The caps weight together 98 gram. Aprox. 55 gram for the fibers and 43 for the epoxy. The caps are 18mm wide.

The core has foam inside and sides frome 4mm balse. Vertical grain, a rather light quality.

The spar is wrapped with 2 layers 58gram/m^2 glas. The famous three layers at the joiners.

So, nothing that looks not good or is not specified. As far as I can see of course.

The questions:

Where did I go wrong? I would learn from the mistakes I make so I can collect expirience and avoide in the future this kind of chrashes.

It is not allowed/calculated to do monsterhigh starts with electrical winches?

If I’m a fool with High starting, Let me know. I won’t be angree at those who dares to say soo ;)

Is it possible to determine the weakest part in the spar by looking at the fraction to specific signs? A local dry spot or what ever.

The left side of the spar is “still going strong”. Is it possible to sacrifice the spar to sience and do a destructive test?
For example: The piece of the rightside of the spar can be hold by clamps or wathever. The left side of the spar is hanging free. The load that can be carried by the joiner should be 41.75Kg (2900lb-in = 33.411 kg-m. 33.411 / 0.8m = 41.75kg)

Wen doing the test and carefully watching the spar. How can I recognice the weakest parts?

At least: In the most extreme case senario I’m ready to send the spar to someone with understanding of the “royal art of spar building”. Maybe he can take a look.

Well, the good side of this crash is that I know now with parts are strong enough. The boom and pylon area don’t have any defects. Just one fraction. The rudder, stab and tibpanels don’t have even a scratch. The radio equipment is also oke. Only a new fuse, centrepanel and a V for the mount.

[markdrela]

During a heavy highstart the spar of the centre panel broke.
Ouch. Major bummer. :(


The sparcaps are built up from layers HM carbon. The glasthread that holds the fibers together is still at the carbon. The caps weight together 98 gram. Aprox. 55 gram for the fibers and 43 for the epoxy. The caps are 18mm wide.
You should take a very close look at the sparcap fracture. Two possible reasons it might have failed at such a light load:
1) Dry spot in the layup.
2) Kinked carbon fibers, possibly from the cross threads.

I experienced 2) about fifteen years ago on the HP hydrofoil project. To make the solid-carbon wings we got some uni carbon which was held together with rather coarse glass cross-threads which were woven across the carbon and bonded to it somehow, i.e. the cross threads were not simply bonded on the carbon's surface. When this stuff was laid up, the carbon tows would have a severe kink at each glass cross-thread, even with very strong compaction. It looked rather bad, so I did a simple bending test of this material. The carbon reliably failed with a clean fracture at one of the glass cross-threads -- at less than 10% of the expected load!!! :eek:
I don't know who thought up that material, but it was unbelievably awful. Anyway, the problem disappeared when we switched to smoother Uniweb carbon material.

Wen doing the test and carefully watching the spar. How can I recognice the weakest parts?
I subjected my spar to a bending test before putting it into the wing. That gives great confidence in its strength.

[markdrela]

PS.
If the cause of the failure is one local dry spot, it may be easier to fix the panel rather than make a new one. You can put in two vertical carbon "blades", glued to the front and the rear shear web of the spar at the break. Each blade should be about 15 cm long, full airfoil depth, and 2x the thickness of the local sparcap. The ends must taper off to zero thickness to avoid a stress concentration. If these are solidly glued to the shear webs, then they will transfer the bending loads very well.

But you should first do a 3-point or 4-point bending load test on the opposite side of the center panel, to see if there are any weak spots there.

[Bryan Quick]

To test the spar before installing it, where do you suggest supporting it and applying the load? How much load?

Thanks

[markdrela]

To test the spar before installing it, where do you suggest supporting it and applying the load? How much load?
To test the center panel spar:
1) Support it close to each end.
2) Push down with 85% of the target design winch load.

For TD, a 128 lb load (simulating a 150 lb winch load) is probably conservative. Load more for F3J. Standing on a bathroom scale and pushing down is a simple load method.

A 170 lb load (simulating a 200 lb winch load) will make it bend quite a bit.
But it will be far from failure if there are no defects.

PS.
The above test properly simulates the spar stresses in the center. But it underloads the outer parts of the spar. To test the spar halfway out, you can do a half-span test:
1) Support spar at one end and at the spar center.
2) Midway between the supports, push down with 90% of the target design winch load.
3) Repeat on the other side of the spar.

If you have a foam-core spar, you have to be careful to spread out the applied load a bit for this half-span test. There is no "hard point" in the spar there, so it's easier to dent.

[Bryan Quick]

What about supporting on the joiner rods? Would there be any reason not to?

[Christian Baron]

Berrie,
you wrote the spar broke 330mm from the center. Is it exactly the place where you have the flap servo cut out?
It normally has nothing to do to carry the bending forces, because this will fully be done from the spar, but maybe you have had some torsonal loads during the start of your Supra by using the aileron. A combination of bending load and torsion with a notch at the servo bay as stress point with factor 3 higher stresses than normal, can be the reason of your spar failure.
You have used the very sensitive HM UD-tapes. They are very sensitive for fibre breaks during lamination. Maybe there was a weak point too.
Christian

[Berrie]

Mark, thank you for the response.

The hydrofoil project, was it this one?:

http://video.google.nl/videosearch?hl=nl&q=mark%20drela%20record&safe=active&um=1&ie=UTF-8&sa=N&tab=wv#

Its funny, about the same time back I build a hydrofoil when I was a teenager.

Back on topic.

The fracture is not located near the flap servo as you can see at the picture, Christian.

I seperated the broken pieces an take a look at the fracture. The vertical grained balsa is crushed and the carbon looks like brittle rubbush. I don't know if that is an indication what the weakest part was.

I will test the spar thats left this weekend.

I'm not gonna repair this panel. You suggestion is nicely, but no. I want a Perfect Supra ;) Foamcores are already cut.

Now I need to make a decision what to do for the new one. I am strongly thinking of using the normal carbonrovings. This will surely flex more, but better a "swingwing" than a "chrackwing". As I learn now using HM carbon has a lot drawbacks. Christian mentioned a point I never heard about. :eek:

And make the spar a little wider this time. That 20-30 gram won't harm the flying weight if it makes me feel more leisurely. (if thats the right word the online translation gives to me)

Cristian has a nice rovingwettingmachine on his site. That might be a nice thing to try.

I also gonna test the spar before installing him. Wish I had done that the first time.....

The test methode is easy, Mark, but for the 170LB test I need to carry a backpack with some PB-accu's! :rolleyes:

[markdrela]

I seperated the broken pieces an take a look at the fracture. The vertical grained balsa is crushed and the carbon looks like brittle rubbush. I don't know if that is an indication what the weakest part was.
From the way the layers separated, it looks like the material was not fully saturated.

One good way to ensure full wetting of multiple uni carbon layers is to always wet each layer from the bottom only. Put down an "excessive" layer of epoxy, then place the dry carbon on the epoxy, and press down with a dry roller or brush. If there are dry spots, then lift the carbon and apply more epoxy underneath.

Once the carbon becomes all wet, then you can roll down firmly with the wet roller, and add more epoxy for the next layer. If you use the wet roller right away, you can easily trap a bubble of air which will result in a dry spot.

Our solid-carbon hydrofoil wings had up to 35 layers of uni carbon in each mold half, and any dry spot would almost certainly have caused a failure and destroyed the vehicle. The bottom-wetting technique was essential to guarantee complete saturation.

An alternative method is to wet each layer separately, and then stack all the wet layers. This works OK, but is more time consuming, and can mess up the fibers more.

[Berrie]

An alternative method is to wet each layer separately, and then stack all the wet layers. This works OK, but is more time consuming, and can mess up the fibers more.

Thats what I have done. But.....

If there are dry spots, then lift the carbon and apply more epoxy underneath.

I feld it very hard to detect which places are dry. I had the expirience that the carbon looks for a few seconds shiny during wetting and after that the shine diseapered, and that makes uncertain of the wetting quality. Is there a trick to detect dry places punctual?

[Christian Baron]

Berrie,
forget my assumtions with the combined loads and the servo bay.

To break the spar at this place were it has happened, there must have been something wrong with wetting or displacement of fibers. HM und UHM fibres are very sensitive to fibre cracking during lamination. HT fibers are more forgiving and therefore I would recomment them for your new panel.

When you look to the broken spar cap of the top site, do you have fibres there that are looking and feeling like fresh unsaturated fibres?

To be shure to have all spar UD-layers fully wetted, put them for some minutes between two PE-Film layers and vac them. Besite Marks good remarks for weeting carbon UD-layers this can help too to be on the safe site. I'm using UD-layers only in moulds and with vaccuum. I've build the spar for my extended Supra wing with carbon tows with my simple wetting machine.
Christian

[Berrie]

I did the test. Not good. I could push with +/- 20Kg at the end of the spar (halfway out, the short broken piece was hold down between the desk and a piece of wood and clamps). Then, a "dry cnack", and the spar broke. This time the fracure was (almost) next to the wood inside the spar, close to the bold beam. That could be the stressriser.

No more HM carbon for this boy. I think I underestimate this material.

[markdrela]

I can clearly see the white cross thread at the fracture.
Is this cross thread woven into the carbon? (this is the worst situation).
If the cross thread is merely on the surface, is it relatively thick? (this is also bad).

Ideally, any cross threads present must be on the surface only, and must be thin and/or soft enough to squash very flat under vacuum.

[Christian Baron]

The white cross thread is only on the surface of the unidirectional tapes. They are only on one site and they are out of a soft polyamide.

In my opinion, the very low compression properties of high and ultra high modulus carbon fibres are the key factor for this very brittle break.
In F3B wings you can buy, they are using these UHM an HM Fibres for the bottom spar cap only. The top spar is produced with HT fibre unidirectional tapes and all with a vaccuum consolidation.
I think kinking of these high modulus fibres in combination with disorientations through the cross threads and the very low strain at break are the reason for the bad behaviour of these fibres in hand layed laminates.
HT fibre unidirectional tapes are by far not so sensitive.
Mark's Supra with the use of premanufactured spar caps is the best way to get all the properties out of the carbon fibres thats possible. Fibre alignment is perfect und when you get these ready to use material with HM fibres, you will also get better properties in compression wwith them.
Christian

[Berrie]

A picture of the suspect white cross tread. It feels decently hard under my fingertips. The layer thicknes with this fibers is just 0.07mm. So if you want at least 0.5mm at the spar end, there are 7 of these crossthreads. Sparcentre thickness 2mm gives 28 crossthreads. :eek:

The next moves in the chessgame:

* Use Mark's wetting techniq
* Find a way to fabricate easy a capstrip under vacuum.
* Make a pair of capstrips from HM and HT carbon
* Subject them to a 3-point bending test with the hope that there is a concurrency between the expected qualitys of HM and HT roving produced with hand lay-up. (thanks Christian, for your input)

Any additional ideas for a "highquality hobbyshop's" test?

(some people advised me to stop trying and buy prefabricated capstrips. Is it really so difficult to make a proper capstrip? I wont be pigheaded, I just love to make my own stuff. If I could, I would plant a carbontree and bake my own carbon from the carbontrees hull...)

And, a question to the group, did anyone ever tested his spar the way Mark desicribes (including the loads) or do you all say a prayer and shoot the plane into the sky?

[ViktorF]

Hi Berrie,
Sorry about lost your center panel.
Ours Supra's spars still alive despite of hard launches in a wind up to 9-12 m/s.
Not tried full pedal winch yet, hi start only.
Caps are not pre tested properly. Just bending in the hands.
We used ~0.1 mm "Russian" unicarbon and L285 epoxy.
- foam rolling the plastic film with epoxy
- laying 250mm wide carbon
-covering with other piece of film
- hard rolling
-cutting strips
-remowal of a film
-laying and clamping
Middle panel caps was 20mm wide and a bit thicker.
Fully assembled spars had 10-20% of weight penalty.
Hope this helps.

[ViktorF]

One more thing...
One time the friend was gone with his Supra too far and low and has lost the radio control.
The glider has hit the ground vertically from 10 meters of height.
Damages only at the fuse and middle part of a wing.
Smashed fuse front and nosecone, one wing servo dead, wrinkles and delam on a wing top, broken the double(!) kevlar skin on a wing bottom. But the spar has survived.
We have replaced a front part of a fuse, rebagged the middle panel, repainted repaired areas.
His Supra flyable again.

[VHO]

After some off line discussion, Berrie encouraged me to post how I make spars here.

The method I use was originally described to me by Alan Mayhew, but I have adopted it in my own inimecable way. So all responsibility for any failures in use are mine. So far so good. Have launched hard enough to have significant wing bend, even done a wing over under winch power (for those of you who remember control line that needs no further explanation) and although I have broken the bolt beam the spar emerged unscathed.

Materials.
Ingredients. 300g/m^2 UD carbon, West Systems 105 resin and 206 hardener. PVA release agent.
Environment. Building board with preset 5deg difference (2.5deg each side). 3mm Al bar 25mm wide, bent to dihedral, painted with PVA release agent. Peel Ply (some form of nylon taffeta like material), dark felt, Vaccuum bag, and weights.

Method.
1. Cut carbon to suitable strips. I Use slightly wider than 20mm, and because the dry thickness is approx 0.28mm, use 9 strips, 3 full length (1600mm if building a spec Supra, in my case 1500mm), 2 1200mm, 2 800mm and 2 400mm.
2. Layer the strips in order, 1600,1200,800,400,1600,400,800,1200,1600 as follows.
3. Wet out the bottom side of UD C strip (opposite to plastic backing). Lay wet side down on PVA release agent coated Al bar, peel off plastic backing wetting top with resin as you go.
4. Repeat process until all strips layed up on top of each other. Simply put, the bottom of each UD strip is wet out, then the top when placed on bar, then the next strip is wet on bottom, placed on top of layer on bar, plastic peeled off, wet and the process repeated.
5. When all layers are in place, cover in peel ply, felt, put in bag and pull down as hard as your pump will go (I use 24-25" Hg).
6. When dry (next day), remove from bag, take off felt and peel ply and trim to width.

This makes a spar of approx 70g of which approx 22-24ml (20-21g) is resin (approx 30% w/w resin). I mix 2 batches of resin during the layup (10 ml 105 + 2 ml 206). If the resin looks like its thickening, mix another batch. Thinner is better.

I deliberately make thicker (approx 20%) and wider (approx 10%) spars than specifications knowing that I cannot make spars to the high standards of prebuilt spars. Finished weight in a "normal" Supra using old Vladimir Fuselage and boom is 1700g. Anticipated mass of current build (3.6m span, thicker root wing, lighter fuselage) is 1750g.

Lots of pictures attached, some may be instructive, a few are just dreaming.

Hutton

[Berrie]

Thanks Victor and Hutton for your explanations.

Before I return to building the spar I would like to discuss about the results of the test I did.

I build to testpieces with a lengt of 400mm.

Data:
HM carbon weight 12.2 gram, epoxy weight 5.5 gram. (34.7% epoxy, 65.3 % carbon)
Dimensions (at fracture): 21.35mm x 1.1mm

HT carbon weigt 11 gram, epoxyweight 3.5 gram. (24.2% epoxy, 75.8% carbon)
Dimensions (no fracture): 21.15mm x 1.05mm.

As you can see the amount of epoxy at the HT spar is to little. That gives interresting thinks to see during the test. Even at vacuumbagging the caps you can see the difference between the epoxyamounts in the HM and HT spar. The HM spar is wetted with Mark’s methode. Wetting rovings that way is more difficult, or, take more time, care and epoxy to wet out. However, I like working with roving more! It was faster to work with, therefrom the bad result…

The testpreparation: the distance between the imposition is 300mm. The load is located in the middle. The roughsite was upper. (pressureload during the test)

The load was affixed by a specialst. A small bucket was carefully filled with sand till the spar broke or…..

The results:

The HT cap did not broke. It just slips between the applypoints. The load at that moment was:

1930gram.

During raising the load I was carefully looking at the pressuresite of the cap. As you can see, there are dry spots. Look at the SMALL delamination in the front of the picture. That appears under pressure load! The long indication could already be seen before testing. It is a dry spot. The ammount of epoxy, or at least the distribution is not good.

The HM cap broke. Look at the fracture. See is EXACTLY located at the angle of the crosstreads! What a beauty! The fracture looks clean and sharp. Far more sharp than the fracture at the accursed spar that broke under winchload. That fracture shows much more parings. That could be because this spar was wetted properly?

The load was: 2880gram. Sand alone was not enough, a PB-accu was first put into the bucket. Rest filled with sand.

The questions:

· How could I test the HT cap till it brake and compare the results with the HM-cap test.
· Is the load that the HM cap could bear high or low?
· If taking a look at the max. load at the HM cap, the fracture located at the crossthread is a DON’T GO. I mean, don’t use HM carbon again?!
· Or does this test say nothing?! Because the caps are not tested at pressure only.
. Did I make the wrong choise when I bought the HM UD carbon. Should I try ROVING instead of the UD fibers with the crossthread. (as Mark mentioned in one of the first posts............)

[Jurgen]

Hello Berrie, I thought you're bigger :)
Nice you have the courage to continue, congrats !

In analogy to pre tensioned metal in reinforced concrete (aren't they building bridges that way??), should it be a good idea to pre tension the roovings during the curing? The pre tension might also straighten out the fibres but more important is: would there be a use for pre tensioned built spars? Just a question that pops up.

Second, would it be a impossible nasty work to remove the white strands?
Cheers, Jurgen.

PS: for saturating, did you already use a heat gun, or hair dryer? It turns the epoxy into "water fluid".

[Berrie]

Hello Jurgen, I am bigger! My son, Laurens (4), likes to help me. It is also almost impossible to do serious building when the kids are walking arround. I have 3 of them (1, 4 and 5 years age). It is good for the fameliemood if the kids could help. And, it his HIS equipment what we are using for this test, so I had no choise..... When I finished a spar the kids may test it. :) If the spar survives that it will be strong enough for launching to the moon...... :rolleyes:

I did not use a heater to liquify the epoxy. Never tried that. Has it no backdraws? for example, less time left over for vacuum out the epoxy thats to much?

I have never heard of pretensioning the fibers. (only the bend thats in the centre of the original Supra sparcap might be called "pretensioned", if I'm right)

Steel is a very different kind of material than carbon! Steel is, in the principle, more or less a kind of chewinggum.

[Jurgen]

Yes Berrie, heat will shorten the cure time, but I guess the duration of the heat will be mandatory in this case. Maybe try it on the next experiment (or just a little piece to see how the heat affect the saturation), the epoxy really runs into the fibers like mad when warmed up. But too much heat might do bad, perhaps. Just gently heating the spot you're working on, the effect shows pretty fast. If this is bad advice for strength of the epoxy, please chime in.

Talking about heat, one can also make a temper box (out of Styrofoam plates) with a light bulb inside and a ventilator, for longer periods of elevated temperature.

Nice job done Laurens! Nothing funnier than destructive tests isn't it :)

The pre tension was just an exotic idea with supposed potential.
Keep up the nice work! Cheers, we drinken er eentje op :). Jurgen.

[Jurgen]

Just to be sure, it's heating shortly the spot you're working on, not the epoxy container. J.

[markdrela]

The HM cap broke. Look at the fracture. See is EXACTLY located at the angle of the crosstreads! What a beauty! The fracture looks clean and sharp. I think you've solved the mystery. I'm surprised that more people haven't experienced such premature failures with this material.

Far more sharp than the fracture at the accursed spar that broke under winchload. That fracture shows much more parings. That could be because this spar was wetted properly? The many "flakes" at the spar failure indicates poor bonding between the carbon layers. This can be cause by insufficient epoxy, or insufficient compaction, or both.

How could I test the HT cap till it brake and compare the results with the HM-cap test. Use a shorter span between the supports, to decrease the bending deflection at failure.

Is the load that the HM cap could bear high or low?
At your failure load, I calculate a stress of 72 ksi (500 MPa), which is pretty low -- comparable to what 7075 aluminum can take. The carbon should be able to take at least 3x more load. This is still much more than the stress in your wing spar at failure, so something else was going on there (see comment above).

If taking a look at the max. load at the HM cap, the fracture located at the crossthread is a DON’T GO. I mean, don’t use HM carbon again?! Or does this test say nothing?! Because the caps are not tested at pressure only. Did I make the wrong choise when I bought the HM UD carbon. Should I try ROVING instead of the UD fibers with the crossthread. (as Mark mentioned in one of the first posts............) The main problem is the cross-threads. A secondary problem seems to be insufficient compaction. Using HM tow with good vacuum bagging should work fine.
It's also good to try to get the tow as straight as possible -- maybe strongly pull on the ends after wetting.

[VHO]

Just to finish the spar cap build story. Spar caps are out of the bag. Pictures tell story.

Just a few additional thoughts.

1. Don't make the UD Carbon layers end exactly over the top of each other (+/- 20mm is good enough)
2. With heavier UD to start with there are only 8 interfaces with the nylon (?) cross threads. This might be a stress riser, but evidence to date is there is no problem with the UD I am using.
3. Spar caps are bendy, but spar is not (as much)....

Hope that helps Berrie.

Hutton

[nuevo]

mark said it better than I could. I'd get rid of every one of those cross threads.

[markdrela]

Berrie,
I looked at the "De bouw van een Supra" thread at www.modelbouwforum.nl , noticed that you didn't use a vacuum bag to cure the sparcaps (posts 61-66). Instead, you used a piece of lumber and a bunch of clamps. It's possible that this didn't provide a uniform or a sufficient compaction pressure, and made cross-thread problem worse.

When/if you make another spar, you really should use vacuum. Heating the layup is also a good idea. For the DLG booms that I made out of HM tow, I used vacuum and a 160F (70C) cure --- they came out amazingly well. Even 50C is better than room temperature.

[jirvin_4505]

..snip...

When/if you make another spar, you really should use vacuum. Heating the layup is also a good idea. For the DLG booms that I made out of HM tow, I used vacuum and a 160F (70C) cure --- they came out amazingly well. Even 50C is better than room temperature.

Mark a good point.

My only experience with demanding uni carbon work has been in DLG booms and molded wing spars.

In the DLG booms I haven't found a way of removing the pesky threads.

Haven't woried about it in the molded wings as they aren't breaking:eek:
Lots of molded wings out there with the threads still on

The use of the vacuum plus hairdryer and rolling certainly gives a good compaction with low resin %.

cheers jeff

[nuevo]

jeff, thanks for sharing your wisdom and experience. I see my comments were misguided.

[jirvin_4505]

jeff, thanks for sharing your wisdom and experience. I see my comments were misguided.

Maybe not;)

Had a little play in the shed. The strands can be individually picked off but it is a messy operation that can leave the carbon disarranged:eek:

Many years ago when the only uni availble had thick paper cross web we would comb the carbon to remove the webbing - this may have some application.

My gut reaction tells me anything mechanical is going to do more damage than we are trying to avoid - especially with the brittle UH carbon.

Remembering Mark's posts on desizing carbon tows using a solvent I just had a little go at the threads with paper towwelling soaked in Methylated spirits. The threads were much easier to remove, wiping away with minimal mechanical pressure.

So now I am wondering if the threads become loose and thus able to be removed when saturated with epoxy??

Early days yet:confused:

cheers jeff

[jirvin_4505]

Thread removal ...Tested with epoxy this didn't loosen the threads the same way that the methylated spirits did. Very messy operation trying to remove the threads once wet out with epoxy.

cheers Jeff

[Berrie]

noticed that you didn't use a vacuum bag to cure the sparcaps (posts 61-66).

You are completely right. That was a major blunder. It makes me feel: :o

I will warn in that posts on modelbouwforum not to do what I have done.

(See also post 111 for the centrespar)

When/if you make another spar, you really should use vacuum.

I will. I'm just thinking how to make a nice "mold" that also can be used for the tipspars and other planes. I want to avoid trimming after curing. I have some ideas. The ideas need to "ripen", takes time.

Heating the layup is also a good idea.

A moment ago I give that a try while making the new V, for the mount. A little hard to see on such a small part, but I think it works! I dapped with a tissue. Before heating, the tissue didn't suck epoxy. After heating it does!
So I will use that trick.

If making a sparcap, what is te right moment to heat the wetted carbon? Halfway during wetting the spar, or in the end? Just before it goes into the vaccuumbag?

I think VHO has a good point with mixing the epoxy in two batches.

Remembering Mark's posts on desizing carbon tows using a solvent I just had a little go at the threads with paper towwelling soaked in Methylated spirits. The threads were much easier to remove, wiping away with minimal mechanical pressure.

Another good hint!! Mark, I'm waiting for your book ;)

I also tried that half an hour ago. It works! But I doubt it is an serious option for making a hughe sparcap as the centrespar caps are. If you also take a look at the brittle and fragile HM rovings I lost my enthousiasm for HM spars. I'm also pretty sure that a homemade spar, with equal cross-cut, is not as strong as an prefrabricated one. So you need to make a larger spar, as Victor and VHO succesfully did. If making the spar cross-cut larger, the spar becomes automatically stiffer. Right? So the requisite for a stiffer fiber is not so strong anymore. Even if my new spar is about 100gram more heavy than the old one, the flying weigt will be below 1500gram. I can live with that. So my choice is: HT roving 1610 tex. I buy me 2KG, enough for about 10 Supra's :)

Remember I have still HM spars in the tip panels......... :rolleyes:

[jirvin_4505]

Berrie any updates on your spar building.

Looking at your tests again on the short pieces - these were vacuum bagged.

Was any heat used to help the wet out and cure?

Have you tested the HT spar on narrow supports to see the failue mode?

cheers jeff

[Berrie]

Hey Jeff, good to notice that you are back!

Thanks for your interest in any updates.

Looking at your tests again on the short pieces - these were vacuum bagged. Was any heat used to help the wet out and cure?

No, I did not use any heat.


Have you tested the HT spar on narrow supports to see the failue mode?

No, not yet. Mainly because of the fact I didn't wet out the spar good enough. So I do not expect representative values. But I will give it a try this week. It remains funny to do destructive test anyway....
(I will keep the best wetted site of the testcap on the pressuresite during test.)

Berrie any updates on your spar building.


Well, now you have asked it, yes, there are a few updates. Not completely tested, but I will share them.

First of all, I have "calculated" the amount of 24K 1610tex HT rovings that Í want to use for the new capstrips. With 10 rovings at the end of the cap and 36 at the centre (for one capstrip) I have more or less the size of the capstrips Hutton and Victor use.

I'm also almost ready with the mold. I did this weekend a first dry testrun. I think this should work. The mold is adjustable, cause I want to build other gliders to. I can make sparcaps straight with a lenght of 1600mm and a wide of 40mm by 5mm thick (for the case I m gonna build that Flying Special 5000 clone.......) Tapered capstrips are also possible, from 40mm wide to zero mm wide. It can taper in two directions so capstrips for a complete Supra wing tip can be made in one session. If I want to make a capstrip with dihidral, the maximum lengt is also 1600mm. (800mm to the left and 800mm to the right)

The mold consists of 6 pieces wood. Two of them are "groundplates" with bolds. The groundplates are connected, very tight, with hinges. The other four have slotted holes. The nuts are covered with something I don't know the englisch word for it. Its stuff you can use to cover the kitchen or toiletfloor or wathever with. This protects the bag under full vacuum.

The mold is covered with cheap wide tape to prevent the capstrip from sticking to the mold. A shaped icecreamwood is used to cover into the corners. (my kids trow these things into the dishwasher instead of the garbage can)

The piece of steel in the centre keeps the dihidrel more or less. It helps while I'm putting the hole construction into the vacuumbag.

The pictures are more clear than my words.

[Christian Baron]

Sorry!

[Christian Baron]

I wrote some threads before, that compression properties for UHM fibres are significantly lower than tensile properties and also much lower than values for HT-fibres.
Here are measured properties on UD laminates:
..................................HT.............. UHM
tensile strength:.........2050 MPa......1921 MPa
compression strength:.1985 MPA.......1006 MPa

The properties for HT-fibres are nearly the same in tensile and compression.
The properties for UHM-fibres are only ~50% in compression compared to the tensile value.

Christian

[Erik.R]

First of all, I have "calculated" the amount of 24K 1610tex HT rovings that Í want to use for the new capstrips. With 10 rovings at the end of the cap and 36 at the centre (for one capstrip) I have more or less the size of the capstrips Hutton and Victor use.


Hi, I'm going to start a Supra build this winter, I've been planing on making the spar with UD in an aluminum "U" channel, luckily I found this thread and learned from your mistakes :rolleyes:
I've done a test with the U channel It is very difficult to get the spar out without breaking is, so I think your on the right track with a built up U channel.

I have also calculated the quantity of 24k to build the spar with 20% added for safety, this is what I'm planing on doing:

2.7mm>0.6mm => 35tow>8tow

8x 1600mm
4x 1400
4x 1200
4x 1000
4x 800
4x 600
4x 400
4x 200

So that's about 72m of 24k tow for both central spars, I guess I'd better build a wetting system for the tow :eek:

Keep us updated with the build :)

Erik

[Berrie]

Hello Erik, very wise, learning from someone else his mistakes! I wish you al the luck and joy while building your Supra.

I did the test with the HT testpiece. It was stunning!

Laurens is sleeping so I must use my own equipment.

The first test:

distance 200mm, load 5076gram. It didn't broke. It just slipped between the supports.

The second test.

Distance 150mm, load 8591 gram. It didn't broke again!! I just heard some cracking, and it slips again between the supports!! The amount of energy that I feld while releasing the load was scary!

If I'm not mistaken, this badly saturated capstrip, is at least 50% stronger than the HM cap. (just looking at the moments)


Thanks for pushing me to do the test, Jeff.

This test proves to me the warnings that came from Christian. Btw, thanks for sharing the information in your last post. It gives me also a comfortable feeling about the decision to use HT carbon. I ordered this evening carbon and glass (especialy for Hutton ;) )

[VHO]

Hello Berrie,

Thanks for you thoughts on a mould/jig for building spars.

Made a bunch of straight spars (for tips) with just 20k(ish) rovings using the lay up procedure I described earlier.

It is not so good maintaining the external dimensions, as with strips cut from UD held by webbing. Can see that a place to lay the individual wetted out rovings, and better still to constrain them somewhere near final position while you ensure they really are wetted is a good idea. Your mould idea gets a BIG tick in that regard.

But, I am just a little concerned that you might not get even pressure over the entire width of the spar especially when it gets down to 0.05mm (or less) thickness, and that the section might be thicker at the edges than in the middle and have high spots on the edges (where you will be laying glass wrap over to build the spar). Have you tried it out to see that this is not an issue?


Hutton

[Berrie]

Can see that a place to lay the individual wetted out rovings, and better still to constrain them somewhere near final position while you ensure they really are wetted is a good idea. Your mould idea gets a BIG tick in that regard.

I don't really understand what you are saying.... Can you clerify?



But, I am just a little concerned that you might not get even pressure over the entire width of the spar especially when it gets down to 0.05mm (or less) thickness, and that the section might be thicker at the edges than in the middle and have high spots on the edges (where you will be laying glass wrap over to build the spar). Have you tried it out to see that this is not an issue?

You got it right! That is the only backdraw that I have to deal with. But if I carefully place the vacuumbag over the gap where the capstrip is in, the problem is solved.

And even if the edges are a little ticker, with a file it is just a few minutes work to remove. The testpieces are made in the same kind of molds.

[VHO]

[QUOTE=Berrie]I don't really understand what you are saying.... Can you clerify?

Its just that wetting out and laying up of rovings requires a bit of "poking about" during the layup. When they are in a "well" they will be less prone to escaping, and this should make both the wetting out and alignment of fibre direction easier to do.

Might make something similar and see if it helps.

Keep up the good work.

Hutton

[Berrie]

Oke, thanks, now I understand you.

Keep up the good work.

Succes consists of going from failure to failure without loss of enthusiasm :D

I'm now thinking of building a heating box to. Yesterday I did another test with the hairdryer on a part for my vintage F3B glider (a protectionbar under the rudder, since I'm good at stripping servogears) The heat makes the epoxy so noticable thin. Gives me a very good feeling.

[jirvin_4505]

...

Its just that wetting out and laying up of rovings requires a bit of "poking about" during the layup. ..snip..

Hutton

Roving wetting machine and peg board for layout - from one of my all time favourite threads - thanks to M Koch

http://www.rcgroups.com/forums/showthread.php?p=1229240#post1229240

more pics...
http://www.rcgroups.com/forums/showthread.php?p=1250514#post1250514

cheers Jeff

[jirvin_4505]

Anybody had experience with this uni carbon from soller?
http://www.solarcomposites.com/composites/carbon%20fiber%20uni-fabric.html

http://www.solarcomposites.com/images/UNIback.JPG
Back of UNI (notice the tiny strips of hot-melt binder)


Specs: 9.0oz. sq yard/305gsm Super High Quality, T700 High Modulus, 711Ksi, 34Msi, 0.014" Thickness, 15.9ft/lb Carbon Fiber Uni-directional fabric

Specs: 4.1oz. sq yard/139gsm Super High Quality, T700 High Modulus, 711Ksi, 34Msi, 0.006" Thickness, 35.2ft/lb Carbon Fiber Uni-directional fabric


Is T700 High Modulus the brittle stuff?
Is the hot melt binder a problem?

cheers jeff

[markdrela]

Is T700 High Modulus the brittle stuff?
Is the hot melt binder a problem?
Compared with "normal" carbon like T300, the T700 fiber has
about the same modulus, but about 2x greater strength.
So it will be able to deform about 2x more before failure.
I don't know the effect of the binder strips are.

[Christian Baron]

Here you can find the product data sheets for the Toray carbon fibres:

http://www.toraycfa.com/highmodulus.html

The link is going to the high modulus fibres, but under product information you will find the data for the T700.
By the way, in the data sheet you will find the laminate properties for tensile and compression. Compare these data and no surprise you will find properties I have measured too and shown some threads before.
Christian

Jeff, you got my mail?

[jirvin_4505]

Thanks Mark and Christian for reply.

So I could conclude that the use of the T700 fibre is no advantage in a supra spar as they are designed for stiffness and already over strength. (maybe an advantage in a crash!) The only advantage then is price and availability - not that I have investigated price and availability from other sources yet.

Some local feedback on the use of the Soller supplied T700 uni..
http://www.rcgroups.com/forums/showthread.php?p=10662138#post10662138

What do full size manufacturers do about the binders? When designing fullsize do they account for the depredation in performance and calculate for it?

From a Torah's perspective there must have been some research into binder performance and affects in laminates. Striving for high performance carbon fibres and then degrading the performance with binder systems seems counter intuitive?

Am I sweating this binder thing too much?

The Toray's carbon fibers data sheets presents many interesting choices.

The MJ fibre .. stiff and strong

Christian thanks for email.. :)

cheers jeff

[ViktorF]

All these data put me on dark side. I even dont know what sort of UD I use for spars.
Any body can say what is it?

Though ours Supra spars hold high loading so far. See pics at Supra build thread.

[Berrie]

hey Victor, I dont think there is a reason for you to be scared. I think I'm the only one in the world who blow up his homebuild Supra......

This post has two partst. A fun part, the building. And a boring part. Some calculations. I hope Mark will take a look at those, so I could be sure they are right and I can build sparcaps for "happely ever after"

Part one: The last tests before the "real thing":

I build two more testpieces. To test the new mold, work with the hairdryer and do a destructive bending test.

The first test did not tun out very good. As you can see in the pictures it is not possible to make capstrips in this mold without a "backingstrip (Or whatever you might call it.)

The hairdryer does his job very good. It is amazing how the epoxy becomes so liquified.

Sadly the first testpiece came out a kind of U-shaped. The vacuum works mostly in the centreline of the cap. This pushes the wetted rovings to the sidewalls of the mold. Therefore the U-shape. (se picture) I did also use peelply to suck out epoxy. Only 0.5 gram epoxy was sucked into the peelply. I think a sign of bad vacuum distribution.

A destructive bendingtest on this testpiece was useless. Didnt do that.

Now the second testpiece. Again 16x 24K rovings with a lengt of 400mm. Weigt 10.4gram (therefor, 1.625 gram for one meter).

Mold covered again with cheap packingtape. Wetted in the mold with a flat brush (works very quick by the way). Bottom first. Put three rovings next to each other. Each roving 7mm wide. Rolling with hard roller. Again wetting with richtly epoxy. Then the next rovings. Rolling with a hard roller, and so on.

When al the rovings are done, and thats not much work, I took the hairdryer and blow the wetted rovings real warm. The epoxy runs like crazy (who said that before?!?) While heating I run with the hardroller over the rovings. You feel the clinging jelly mass becomes more liquifying.

After that is done, push the rovings from the sidewalls back were they belong.

Then a new mold part, a balsabackingstrip covered with packtape, is put on the rovings. It has a tight fit, but with the epoxy and the slipery tapecovering it fits easy.

(Remember the mold is adjustable. I put the balsa backingstrip between the adjustable mold. Close the moldhalves together without packingtape fix the nuts. Remove the balsasbackingstrip and cover the whole mold with the cheap packingtape. Its al done withing 15 minutes)

Cover the whole mold with some breather, I use paper kitchentowels, and put in in the bag. There is no peelply used this time.

Surprisingly there was more epoxy sucked out the carbon strip as with the testpiece before! The capstrip came out VERY good! At least in my eyes. Only a very little sanding was done at the edges. Exces epoxy and some carbon hairs.

A quick calcultions shows that one wetted roving gives 1.45mm^2. Exactly as Christian wrote a few posts ago.

The final weigt is 14.8 gram.

Part two: The calculations.

1) Firts a check calculation to see how Mark gets the 500N/mm^2 for the HM capstrip at post 58:

HM carbon strip

Wb = 1/6 * b * h^2

b 21,35 mm Capwidth
h 1,1 mm Capthickness

Wb= 4,3 mm^3


Mb = F * L

Massa 1,44 Kg (half the load)
F 14,13 N
L 150 mm (half the distance between the support)

Mb= 2118,96 Nmm


Sigma = Wb/Mb

Sigma= 492,14 Nmm^2


This looks like almost 500 Nmm^2 so I dare to show the rest:

2) The badly wetted HT carbon capstrip:

HT carbon strip

Wb = 1/6 * b * h^2

b 21,15 mm Capwidth
h 1,05 mm Capthickness

Wb= 3,89 mm^3


Mb = F * L

Massa 4,23 Kg (half the load)
F 42,14 N
L 75 mm (half the distance between the support)

Mb= 3160,41 Nmm


Sigma = Wb/Mb

Sigma= 813,22 Nmm^2

Oke, a lot better then the HM carbon capstrip.


3) now whe gonna check what a hairdryer an good vacuum does:

HT carbon strip (hars goed verdeel)

Wb = 1/6 * b * h^2 16rovings 1,450625 mm^2 per roving
16*400=6400, weegt 10,4gram(1,625gr/m)
b 21,1 mm Capwidth
h 1,1 mm Capthickness

Wb= 4,26 mm^3


Mb = F * L

Massa 6,18 Kg (half the load)
F 60,63 N
L 70 mm (half the distance between the support)

Mb= 4243,8 Nmm


Sigma = Wb/Mb

Sigma= 997,3 Nmm^2

THAT is COOL!!! The capstrip slipped between the support without any cracking noise or visible dents at the pressure side!! So the capstrip is even stronger than the 1000 Nmm^2.

4 Now the strengt of the HM capstrips from the piece af spar that was left over from the crash. I did use the calculation as Helmut Quabeck shows at his homepage:

http://www.hq-modellflug.de/theory.htm

This is my result:

Length 740 mm
massa 22 Kg
F 216,04 N
Moment 159869,6 Nmm

b 18 mm capwidth
h 2 mm capthickness
Hi 14,3 mm heigt between the caps
Hu 17,5 mm heigt over the caps
Ix 3652,752 mm^3

Sigma 383 Nmm2


So far so good. If I didn't make mistakes, all the advises from you al together learned me to build capstrips with reliable quality. From 383Nmm^2 to 1000Nmm^2. Small step for mankind, but a giant step for Berrie!! Thank you all!!

[Erik.R]

Thanks Berrie this is very useful for building my supra and other gliders.

[Jurgen]

.... I did also use peelply to suck out epoxy. Only 0.5 gram epoxy was sucked into the peelply....Hi Berrie, just to be sure, the peelply itself is not used to store excess epoxy, that job is for paper towel (or even toilet paper for better absorption) that is put on top of the peelply. The peelply is sort of an interface between the wet fabrics and the absorption agent (the paper towel). Maybe you did know this. Cheers, Jurgen.

[VHO]

Berrie,

Who ever thought that we learn most by doing things right the first time :D

You have demonstrated perfectly a learning process that we can all benefit from.

Well done.

I look forward to seeing your completed Supra (again), you getting the confidence to push it hard and then seeing how you recover from what breaks next. ;)

In the meantime, I will try out a few modifications to a Supra wing and learn what not to do the time after that. I'm sure that a home built Supra / Icon2 thingy might just teach me something else I don't know. :)

[Aside] Once upon a time when I flew F2C we used to joke that we were scale modellers (i.e. we copied each others models, now they just buy them). Its pretty much the same with the Supra club. We all try to build accurately to Marks plans = 1:1 scale models. But its the learnings that come from experimentation around the edges of a successful design that really push the hobby forward.

Hutton

[jirvin_4505]

Berrie, A very good tutorial :)

>hairdryer..... one of my favourite composite tools

>first efforts at bagging spar with resulting curve at edges.... I agree that the second attempt solved the problem. When bagging into a cavity the bagging material needs extra plastic to draw into the corners. Here it is common to pleat or tent the vac plastic so there is extra to draw into the corners. From you photo there doesn't seem to be this allowance. In your .nl thread is mention of stretchalon (stretchy bagging plastic) - this is very good material for drawing into sharp corners.

Again well done and I admire your perseverance

cheers jeff
@VHO...hoping to join the Supra club in 2009

[Berrie]

Thank you for the kind words.

With the knowlage of the strengt and needed stiffnes and of course with an eye on the weight, I will use this rovingschedule (24K HT):

9*1600
3*1350
4*1100
4*750
4*500
4*220

This is 28.7mtr. for one capstrip, apr. 47.5gram. After wetting it should have the weight of about 66-70 gram.

[Berrie]

to push it hard and then seeing how you recover from what breaks next. ;)

You get what you asked........

I did the test as Mark discribes in post 39. Support the spar at both ends, standing on a bathroom scale. I saw the scale running back from 70Kg to 15 kg. Then, again, the new spar broke. :eek: The fraction is exactly located in the centre.

I did say some really ugly words this time. :o :(

Sparcap tension must be about 360N/mm^2 (caps: 21*2mm, height overall 18mm) That can't really be the problem, according to the tests I did with improving the production methods and chancing to HM rovings.

I'm thinking of:

1)Maybe it was wrong, even foolish, stupid and dump (dump and dumper, you know the movie? I feel that way now), to left the space for the boltbeam open.

2) What intrigues me is the long fraction at the glaswrap, close to the sparcaps. (it is wrapped twice with 54grams/m^2 glass at 45°)

[Christian Baron]

1)....................to left the space for the boltbeam open.

Yes!!!

Sorry to say it so hard.

Christian

[Berrie]

Sorry to say it so hard.

Fine to me! I like clear answers.

[VHO]

Is it better to fail on the ground or on top of launch?

In this case I suspect that you are being too hard on the system. No boltbean means you have made a stress point at interface between spar cap and spar. The spar should easily have been able to withstand 70kg at 360N/mm^2.

Next time with Bolt Beam? :D

Hutton

[markdrela]

There will be a large shear stress concentration at the open bolt beam hole. To prevent this the bolt beam must be in place, and there must be a few kevlar or carbon tow wraps around the spar on each side of the bolt beam. The wraps are shown in the Supra wing plan. You might also consider adding a 3rd glass layer over the center 30 cm or so, to make the bolt beam installation and wraps a little less critical.

The long shear failures in the glass are strange. Maybe they are secondary damage after the main failure in the center?

Did you notice exactly what failed first? Shear skin tearing? Sparcap buckling at the hole?

[Berrie]

Is it better to fail on the ground or on top of launch?

Ground.


Next time with Bolt Beam? :D

I like your humour. Hahaha, that makes me smile again.

To be serious again, I think i'm not gonna test the third spar.

There will be a large shear stress concentration at the open bolt beam hole. To prevent this the bolt beam must be in place, and there must be a few kevlar or carbon tow wraps around the spar on each side of the bolt beam. The wraps are shown in the Supra wing plan.

Pfff, that you don't lose your patience with me! Respect.

I did know there are wraps shown on the plan. But actualy I thought they had another purpose. My idea was that they hold the upper sparcap and lower sparcap together. Because the winchforce is lead into the spar by pulling on the boltbeam, the boltbeam is pulling the lower sparcap. To bring the winchpulling force also into the upperspar, you need the kevlar/carbon wraps, so the upper sparcap "feels" the winchforce as well.

Tought that the bias glas is bearing all the shearforces. Never expected that those charming kevlarwraps where importent for shearforces to. (together with the bolt beam.)

Do I need a shrink?


You might also consider adding a 3rd glass layer over the center 30 cm or so, to make the bolt beam installation and wraps a little less critical.

I will. Third spar is already halfway.... (get out of bed very early this morning)

Did you notice exactly what failed first? Shear skin tearing? Sparcap buckling at the hole?

Woo, difficult to say. It happens VERY fast, I noticed no failerus while pushing. What I can reproduce. I used two chairs. About 1500mm between them. (risky point, maybe to much) The spar was at each chair (the part of the chair where you put your butt on, nice soft, so I won't damage the spar by any hard point) with the whole wingjoiner part. The bathroomscale between the chairs (predictable). While pushing the spars I was very consentraded at the bathroom scale. I saw he scale running backwards. Suddenly the spar broke, OR slipped from the left chair and hit the ground left first. Because I'm very sure that I heard a loud "tick" on the left of me. While falling my body speeds up, only a short distance. But maybe enough to break the spar. The spar brakes my speed, while I was already pulling with 55Kg on a spar with the boltbeamhole. Reaction force might have been to high, and whe know the result.

It's just a theory I think about, while I was reflecting during "sleep" the night after.

The long shear failures in the glass are strange. Maybe they are secondary damage after the main failure in the center?

Above theory gives maybe an new point of view on the shair failures.

In the end my wife has the most wise observation. "you are as impatient as your son" (Laurens, sweet boy, really) You should not have done a test like this after another hectic day at work while you where so weary".

[Neil Stainton]

Hi Berrie,

Well done for testing! I am very sorry the spar broke.

The bigger problem is that you don't know if it would have been strong enough of the bolt beam had been in place. Would it be possible to repeat the test on each of the broken spar halfs? I am sure Mark will be able to advise on where to place the load - probably not half way. However the force required will be even higher, and perhaps you need to devise a safer way of loading it!

Regards,

Neil.

[markdrela]

I did know there are wraps shown on the plan. But actualy I thought they had another purpose. My idea was that they hold the upper sparcap and lower sparcap together. Yes, that is the main function. But another function is to hold the sparcaps against the bolt beam and prevent a shear peel failure where the bias glass is cut.

[Prop-er]

Wing testing. You know how they did that in the old days?

http://www.sugarpineaviators.com/Luscombe%20people%20on%20wings%201.jpg

Surely we can do this on models too. Here's a Blade DS wing half under sophisticated quality control:
http://www.dynamic-soaring.de/images/fotos/ds_blade/ds_blade_4010.jpg

Big gliders are tested this way:
http://www.dg-flugzeugbau.de/bruchversuch-d.html

A pretty useless way of testing wings. Here's 24 kg of brick...

http://i305.photobucket.com/albums/nn217/prop-er/IMG_1524.jpg

http://i305.photobucket.com/albums/nn217/prop-er/IMG_1518.jpg

http://i305.photobucket.com/albums/nn217/prop-er/IMG_1519.jpg

The proper way of testing glider wings is shown here:
(Not far off your sand bucket method)

http://www.logo-team.biz/cult/index2.htm
http://www.logo-team.biz/daten/bilder/qualitaet1.jpg

http://www.logo-team.biz/daten/bilder/qualitaet2.jpg

Sadly this great method cannot be applied to the remains of your failed spar.

[Berrie]

The third spar is ready. It has the extra layer of glass over the firts 300mm from the middle of the spar. (total 600mm) Of course the kevlarfibers.

@Neil, because of the glassfractures it wasn't possible to test the halves. If I applied some load, I hear the fracture running along the sparcaps. Pretty funny, but not what you want of course....

You are right about the uncertainty if the spar I build this way is strong enough. I'm missing the boldness to test the 3th spar. But I'm very sure it is much stronger than the first one with the HM sparcaps.

According to the bigger capstrips, the HT carbon, the improved production procedures, the extra glas. And the knowledge a spar without kevlarwrap and boltbeam can still carry a load of more than 50kg...........

Altough Prop-er shows us some inspirational pictures (with HAPPY faces on it) I'm not gonna test this spar. (Btw, I like that Shark....should it be possible to bag such a wing. Mmmm, you make me dream of a next project)

I saw recently some new Supra's began to rise. Maybe someone else might try this at home :)

[Jurgen]

At this state, some cheer girls are applicable
http://legslip.files.wordpress.com/2008/04/browny-cheer-leaders1.jpg
GO BERRIEEEEE !! :D

[Berrie]

That will make me forget my Supra completely. :D

Is that what you want to achive Jurgen? ;)