Carbon Reinforced Lattice Bracing.

[Kim]

I did not want to hijack Jim's gallery topic showing his magnificent and innovative 'Telicoustic' guitar so thought it more polite to start a new topic in the build forum so we can voice our views with vigor on the topic at hand.

OK, so if I am understanding this correctly, the popular method employed by most who lay-up carbon reinforced lattice bracing in their workshop is to string 12k carbon towe (unwoven cord made up of 12000 individual and continuous strands of carbon fiber each being finer in cross section than a small ant's dick) between nails placed in line with the intended braces, and while the carbon towe is in tension, to then saturate that fibre between the nails with epoxy to form a solid carbon strip. At least that is what I recall from David Shramm's excellent input here on the ANZLF (Hope he tunes in for this topic, also hope we did not loose his past input in the recent hack and trash but have not checked)

Anyhow, if I have this right, then I have a problem with the process. The problem I have is that it seem to be a process based upon a belief that product is everything and process is nothing. What I am saying here is that oils ain't oils Sole! (ignore that if you are too young to remember the old Castrol Oil ads of the late 70's early 80's). From what I understand of CF, there is a huge variation that exists between the stiffness to weight ratio of 12k towe laid up with epoxy when compared to oven cured, pultuded, prepreg, unidirectional fiber and that difference is due to the 12k towe having a far greater epoxy to fibre ratio making it much heavier and far less stiff than its pultruded, oven cured counterpart.

Now I am not saying that there is absolutely 'no' improvement in the stiffness to weight ratio of plain wood when a towe lay-up is performed...but, if you consider the added weight of the epoxy into the equation, then this may in fact prove to be the case. What I am saying is that when we look at materials for analysis, there is always a temptation to look at the most persuasive figures that will satisfy our expectations and to ignore, or most likely not even find the rest. If we look at the figures for carbon rod, chances are that the data available will most likely be related to the commercially available pultruded, oven cured variant, simply because the author has a product to sell and a point to make about the stiffness to weight ratio that will allow him/her to do just that. This means that when ever people read about CF, it is those very properties, stiff and lite which gets driven home again and again. So it only stands to reason that an assumption can easily be made by the end user that carbon = stiff and light, period.

So back to the point at hand, CF reinforced lattice bracing...If the towe epoxy set up is in fact much heavier and less stiff than the data so commonly available for the pultruded product would suggest, is there any 'real' gain??? If we take a brace and we cap it with 12k towe and then we weigh it, and then we test it for deflection, and then we make another brace from the same core wood, say red spruce, but this time we use the cube rule and make that brace 'taller' than the first by an amount equal in weight to the first with its epoxy.....which one will actually be stiffer and by how much???

Lets just say for arguments sake that the epoxy towe brace, that is our first brace, is only 'marginally' stiffer than our second brace. Where is the point of diminished return?? When do we stop and say its just not worth the effort to mess with all that extra time, material, cost and the chemical exposure attached to the process for such small return?

I will finish by asking one further question. If it turns out that the towe epoxy system is essential a waste of time, but buying in pultruded, oven cured, unidirectional sheet was a better prospect, what is the gain over a nomex sandwich construction that has the potential to take a top closer to being a true monocoque than anything...ie, assuming that your goal in using CF capping is to build an intricate spiders web of bracing for a paper thin top, how can you hope to compete with the light weight, super strong box section that is a nomex sandwich construction?

Cheers

Kim

[Allen]

My only experience with this has been to use Balsa and Paulonia as the filler between 6K CF tow. This wood in and of itself used as a brace has little chance of being successful. But it does weigh bugger all.

However when I laminate the CF onto it, the increased stiffness is just amazing. The resulting lattice matrix weights in at much less than what a bridge patch and associated braces would on my ukuleles, plus the top is about 1/2 the thickness that I would need to go with conventional bracing.

One thing I learnt by doing a series of instruments one after another about doing the laminations, is that a little epoxy goes a very long way. It's easy to think of it like a water based glue when doing the lay up and imagining that it will somehow dry up, but the fact of the matter is that it bleeds everywhere under vacuum.

The upshot of this I believe is that most of the weight of the top / bracing combination is in the top itself. This is one of the main reasons we look for the stiffest piece of spruce / cedar or whatever we can lay our hands on for the top, in order to take the overall top down thinner. Thus saving weight while preserving the appropriate stiffness.

With the lattice / CF method employing Balsa or Paulonia it takes the initial top stiffness out of the equation, as we can compensate for it.

A further observation of the ukes I've built with this method is that they are all at the 0.8 to .09 mm thick in the lower bout, yet not one shows any deformations or "mapping" of the braces of the top due to string tension or RH swings as conventionally brace tops can.

[graham mcdonald]

A couple of points to consider. You seem to be suggesting that the CF/epoxy tow is only laid on one edge of the brace. This can certainly be done, but I would suggest it is a waste of time. You would certainly add some stiffness, but not nearly as much as using the CF between the soundboard and the brace and then on top of the brace, effectively forming an I beam with the CF.

In the same way adding a vertically laminated piece of CF in the centre of a brace would seem to add a lot of mass, and not be as effective as the I beam style. And then if that brace needs shaping it is a messy and maybe dangerous job with the CF fibers

I have never found getting the epoxy into the CF a problem. Greg Smallman demonstrated it 25 years ago, and while he might well have refined the process, I do it simply by pulling a length of tow from a piece of CF cloth (A weft or a warp or something) and then with thin latex gloves rub epoxy along the length of the CF and squeeze it through. Too much squeezing along the length tends to pull some of the fibers out and make it a bit fuzzy, but it seems to work. I have no idea how many fibers in a piece pulled out of the cloth. That is then laid down in position on the soundboard and the balsa laid over the top and lightly clamped to press the balsa firmly down onto the epoxy. When dry shape the balsa lattice, usually into an overall dome and repeat the CF on top, overlapping the ends to lock it together with the bottom layer. With a little practice is is pretty simple to get enough wetting of the tow with the epoxy, just with fingers, but still have enough epoxy to adhere to the balsa and not gollops of squeezout. I don't think it can ever be a neat operation though.

The main problem with using these braces is getting too much stiffness. A year or so ago I had a 'brilliant' idea. I thought if I carved a mandolin soundboard to an allover 2mm thickness and then added back stiffness in the centre by a small CF/balsa lattice I could get a more responsive soundboard with less mass for the stiffness I wanted. The first one used a 3x3 lattice and sounded horrid. Much too stiff. I made another with a 2x2 lattice with less height in the balsa. Still sounded awful. Both soundboards rang with a metallic ting when tapped. I have since been told by another mandolin builder with a much better scientific education than mine that there is a limit to how low the mass of a soundboard can be for a required stiffness. Past that point you have made a banjo. I am going to pull the back off the second mandolin and redo the bracing even lower and see if I can make the idea work. If I had been smart I would have made the top and back much more easily detachable.

Just some thoughts

[P Bill]

Thanks for this thread . I'v been hearing about cf since joining the forum and know very little about it.

[jeffhigh]

Kim, I think when you look at the structure of the smallman lattice system, there are two aspects I would suggest you consider

1. The regular close spaced lattice drastically reduces the span of the top between supports as well as providing cross grain stiffness which is lacking in a regular torres braced top.
It is this, and the support frame around the lower bout which allows the thickness of the top to be reduced so dramatically. Forces from the bridge can be transferred both by compression in front to the bridge and by tension behind it.
And in a nylon string guitar where trebles are often elusive, lightness is a huge goal.

2. if you are going to use a lattice system assembled from members with halved joints,made of light low stiffness material like Balsa, you need some way of capping them continuously and a thin layer of flexible CF tow does that well and allows for ease of capping in both directions.
Undoubtedly this could be done with spruce instead as some makers have.

I have never played a smallman guitar, but I certainly would like to at some stage.

[Mike Thomas]

Kim, these are good questions. I have often wondered why pultruded unidirectional sheet is not used. It is so much easier to handle than tow and you would probably need less of it, and certainly less epoxy. As Allen says, the stiffness of carbon capped balsa braces, is amazing. And as Graham says, the danger is in making them too stiff. I can't imagine that a nomex sandwich alone would provide anything like this degree of stiffness, unless it were very thick. But it may well have other desirable properties.

I am certainly no expert on this but it seems to me that there are two separate issues here. One is making a stiff and light top so that the instrument sounds as you want it to sound. The second issue is structural integrity, so that the instrument as a whole maintains its shape. As I understand it, makers like Greg Smallman, want their guitars to project, so the top is very thin, around 1mm cedar in Smallman's case, braced with a carbon capped balsa lattice. But this is insufficient to provide structural integrity, so he has a separate, hefty internal bracing system that isn't attached to the top. This is quite unlike a conventional instrument, where the top braces are doing both jobs, and are therefore something of a compromise.

[graham mcdonald]

A thought on the nomex sandwich soundboards. My scientifically minded mandolin building colleague (Dave Cohen in the US) and I were talking about moulding arched, nomex cored, mandolin soundboards a couple of years ago with something like a 1mm skin of spruce inside and out. I think Dave came to the conclusion that you would end up with so much epoxy that wicked up into the honeycomb that that would add too much unnecessary mass from the amount of epoxy needed. Mind you there are those who are making classical guitar soundboard that way, but the info I have seen do suggest the epoxy must be applied very carefully so as not to end up with too much excess. It always seemed like a lot of very fiddly work to do and I lost interest, though the moulded soundboard idea still bubbles to the surface every now and again, but there is only so much time...

From a presentation I was at in the US a few years ago from Brian Burns and Randy Reynolds the rational for the nomex cored classical soundboards was that the nomex allowed a stiffer structure for the same mass. They still needed bracing but were a little more efficient, which in classicals is rather important. I think Brian has some info on his website about the tests they have done on stiffness, Q and all that stuff

cheers

[DarwinStrings]

OK, so if I am understand this correctly, the popular method employed by most who lay-up carbon reinforced lattice bracing in their workshop is to string 12k carbon towe (unwoven cord made up of 12000 individual and continuous strands of carbon fiber each being finer in cross section than and a small ants dick) between nails placed in line with the intended braces, and while the carbon towe is in tension, to then saturate that fibre between the nails with epoxy to form a solid carbon strip. At least that is what I recall from David Shramm's excellent input here on the ANZLF (Hope he tunes in for this topic, also hope we did not loose his past input in the recent hack and trash but have not checked)

I use almost the same method as Graham. That is to wet the tow and lay it direct onto the brace, it is not under tension until the strings act on it. The only thing I did differently was that I dragged the epoxy into the tow with a 6mm cheap pig hair brush that I cut so the bristles are about 4mm long. The tow I use is 2mm wide as it comes out of the mat but the brushing flattens and spreads it to 4 or 5mm, I lay it onto the brace in one direction, brush it onto the lightly pre glued lattice, do the other direction then leave it to cure with no clamping other than the few go sticks that I place on the bridge patch to keep it all flexed into the dish. I used the thicker variety of epoxy.

Lets just say for arguments sake that the epoxy towe brace, that is our first brace, is only 'marginally' stiffer than our second brace. Where is the point of diminished return?? When do we stop and say its just not worth the effort to mess with all that extra time, material, cost and the chemical exposure attached to the process for such small return?

The thought for me was that if I use spruce then I may as well cross brace it cause I know that works and makes a excellent guitar. I weighed the balsa I used against the spruce I would use and it was just shy of one tenth the weight (I didn't weigh the epoxy/tow combination) so I figured I could make the top lighter but gauging how stiff the whole thing would be was a big guess and I reckon I can make it less stiff than I did but that I will only be sure of when I do it again. I don't think it took any extra time other than thinking about it and deciding size and shape as opposed to using a method I already know, the lattice construction is easy, saw a few halving joints with a razor saw into balsa then pop it together and handle it gently till I glue it on to the sound board with a caul and a few bit of tow in between. After that sanded it to a dome (sort of) sanding very lightly, which was fast. It doesn't seem to cost much, balsa and tow were cheap and I often don't get through my epoxy before it goes off as I am not edge striping and swinging doors any more.

Using the pultruded one may lose me a little more weight but after shaping the dome I would have to make all the braces in one direction lower by the thickness of the carbon, do able but more difficult. It might be hard to deal with where the plane of the brace meets the plane of the soundboard. No idea about Nomex the first I ever heard of it was on this forum, does it add much stiffness across the board?

Jim

Life is good when you are amongst the wood

[jeffhigh]

The Nomex is basically a separator, keeping the inner and outer skin apart just like the expanded cardboard honeycomb in a hollow core door.
No real additional strength in itself, but by making the sandwich thicker than the sum of the inner and outer skins, it increases stiffness without adding as much weight as a solid panel of the same thickness.

[kiwigeo]

I did not want to hijack Jim's gallery topic showing his magnificent and innovative 'Telicoustic' guitar so thought it more polite to start a new topic in the build forum so we can voice our views with vigor on the topic at hand.

OK, so if I am understand this correctly, the popular method employed by most who lay-up carbon reinforced lattice bracing in their workshop is to string 12k carbon towe (unwoven cord made up of 12000 individual and continuous strands of carbon fiber each being finer in cross section than and a small ants dick) between nails placed in line with the intended braces, and while the carbon towe is in tension, to then saturate that fibre between the nails with epoxy to form a solid carbon strip. At least that is what I recall from David Shramm's excellent input here on the ANZLF (Hope he tunes in for this topic, also hope we did not loose his past input in the recent hack and trash but have not checked)

Anyhow, if I have this right, then I have a problem with the process. The problem I have is that it seem to be a process based upon a belief that product is everything and process is nothing. What I am saying here is that oils ain't oils Sole! (ignore that if you are too young to remember the old Castrol Oil ads of the late 70's early 80's). From what I understand of CF, there is a huge variation that exists between the stiffness to weight ratio of 12k towe laid up with epoxy when compared to oven cured, pultuded, prepreg, unidirectional fiber and that difference is due to the 12k towe having a far greater epoxy to fibre ratio making it much heavier and far less stiff than its pultruded, oven cured counterpart.

Now I am not saying that there is absolutely 'no' improvement in the stiffness to weight ratio of plain wood when a towe lay-up is performed...but, if you consider the added weight of the epoxy into the equation, then this may in fact prove to be the case. What I am saying is that when we look at materials for analysis, there is always a temptation to look at the most persuasive figures that will satisfy our expectations and to ignore, or most likely not even find the rest. If we look at the figures for carbon rod, chances are that the data available will most likely be related to the commercially available pultruded, oven cured variant, simply because the author has a product to sell and a point to make about the stiffness to weight ratio that will allow him/her to do just that. This means that when ever people read about CF, it is those very properties, stiff and lite which gets driven home again and again. So it only stands to reason that an assumption can easily be made by the end user that carbon = stiff and light, period.

So back to the point at hand, CF reinforced lattice bracing...If the towe epoxy set up is in fact much heavier and less stiff than the data so commonly available for the pultruded product would suggest, is there any 'real' gain??? If we take a brace and we cap it with 12k towe and then we weigh it, and then we test it for deflection, and then we make another brace from the same core wood, say red spruce, but this time we use the cube rule and make that brace 'taller' than the first by an amount equal in weight to the first with its epoxy.....which one will actually be stiffer and by how much???

Lets just say for arguments sake that the epoxy towe brace, that is our first brace, is only 'marginally' stiffer than our second brace. Where is the point of diminished return?? When do we stop and say its just not worth the effort to mess with all that extra time, material, cost and the chemical exposure attached to the process for such small return?

I will finish by asking one further question. If it turns out that the towe epoxy system is essential a waste of time, but buying in pultruded, oven cured, unidirectional sheet was a better prospect, what is the gain over a nomex sandwich construction that has the potential to take a top closer to being a true monocoque than anything...ie, assuming that your goal in using CF capping is to build an intricate spiders web of bracing for a paper thin top, how can you hope to compete with the light weight, super strong box section that is a nomex sandwich construction?

Cheers

Kim

Um can you repeat the question?

[sebastiaan56]

When you stick it on, does it work?

[DarwinStrings]

Martin I had to read it about 4 times with the dictionary out. Thanks though Kim for challenging my brain and expanding my vocabulary.

Cheers Jeff, hollow core doors I have done so relate to what's is going on in that respect. So if you go thicker than your average board you will make it lighter and stiffer or as stiff across the grain as well as longitudinally. You could use a balsa lattice in the same way as the Nomex.

Jim

[Puff]

Kim I

[Kim]

Puff,

I got that, I had no question as to 'why' CF capping is used or what the materials employed do in order to achieve that outcome. My question was, given the extra weight and flex that results from over saturation with epoxy in shop made CF towe brace capping, when compared to factory produced unidirectional, putruded, prepreg CF strips, is it worth the mess and effort?? If the commercially available unidirectional CF strips are in fact lighter and stiffer, would it not be functionally superior and more convenient to form your Ibeams using that product instead of towe??? and...If nomex when correctly glued ( I believe the method best employed is to roll the adhesive onto a very flat surface to form a thin film, and then to lay the nomex onto the film to pickup what is needed via transfer thus avoiding excess glue penetration into the core. My thoughts are to use a hot plate and HHG instead of epoxy...of course) provides such a light weight yet very strong plate when placed between 2 'thin' sheets' of spruce, is this a better way to achieve the out come that is sort by those who currently use CF towe?

(There ya go Marty )

I have more to say in response to some of the great post that have been made in this topic already, but will leave it until I have a bit more time. I will say that when I mentioned earlier that it is my opinion that nomex has the potential to take a guitar top closer to being a true monocoque construction than any other method I have seen, I specifically had that real estate in the lower bout behind the bridge in mind. So I do understand that a nomex top still requires 'some' bracing, but I have seen examples that are completely void of any bracing from just behind the soundhole, where a mid transverse brace had been employed, right back to the tail block.

Gotta be worth a look in to I reckon. Please keep the comment coming, even if they are just ideas and theories because this is how we reveal the truths.

Cheers

Kim

[Kim]

You could use a balsa lattice in the same way as the Nomex. Jim

You could Jim, but nomex is aerospace material. I think it is supper heated phenolic resin that has turned to mostly carbon. They use it in jet aircraft wings and the body work of F1 cars...'super' light because it is essentially ash, but somehow still quite flexible. Once glued between light weight substrates such as carbon fibre sheet or spruce, it forms a very light yet incredibly strong sandwich. If you were to use it for a guitar with minimal bracing as I have described above, I would imaging that 'extremely' rigid sides would be the order of the day in lieu of the X brace or lower bout transverse.

Cheers

Kim

[Puff]

Kim - no smart

[Kim]

Maybe Puff, I can see that too. But with a nomex guitar in your hands you could still wear shoes on your feet, and your teeth may stay straighter and perhaps one would be less likely to develop certain 'urges' for men in canoes who have big mutton chop sideburns ...





















OK.. I can see where the banjo analogy is drawn but, I see that as the challenge of the luthier who walks the path of innovation, that is to find it all at the end of the journey, volume, projection AND tone.

Cheers

Kim

[Puff]

No worries

[jeffhigh]

Kim, some comments about nomex construction.
I will preface by saying I have not personally used nomex and have only researched it briefly.

It is probably most accurate to think of the nomex as a spacing material only, not contributing to the stiffness of the end result directly but increasing the effectiveness of the inside and outside plates by separating them.

Most seem to be using 40 thou inside and outside skins and 60 thou nomex
So you end up with a 140mm thick sandwich with pretty much the stiffness of a 140 thou solid plate but only a little heavier than a 80 thou solid.

This is a big advantage in a classical guitar in particular

In this configuration you could reduce your bracing but not eliminate it.
You could I guess increse the nomex thickness to give you a braceless top but then you would be stuck with a uniform very stiff plate over the whole area.

[Kim]

In this configuration you could reduce your bracing but not eliminate it.
You could I guess increse the nomex thickness to give you a braceless top but then you would be stuck with a uniform very stiff plate over the whole area.

Hey Jeff,

As per my post above, the use of the word "braceless" was specifically related to that area of the lower bout behind the bridge, there was an upper and a mid transverse brace. On my old PC I had images of the same and the plate was not uniform by any means. Rather there was a central 'diaphragm' type configuration shaped somewhat like a pair of fused kidneys. This diaphragm extended out to a solid spruce outer ring or gasket if you like which fixed to the sides allowing it to be feathered in the usual way to introduce more movement into the top as desired.

I have seen similar on the net with minimal bracing constructed in the following way. You start with a top of regular thickness, app 3mm, and then rout out a section of the lower bout to within app 40mm of the outer edges. The nomex is then laid into this rebate with adhesive and, if I recall correctly, the 'nomex' is then reduces to the desired thickness via drum sander before the bottom, or inside .040" sheet of spruce is applied to complete the sandwich. The top I had referred to earlier used this system but had no bracing whatsoever after the middle transverse which sat just south of the soundhole

The point here is that in each of these examples, braced and unbraced in the lower bout the nomex was restricted to that area of the lower bout and I have never seen it extend to the rims. I think doing so would not only be dicey in regard to binding and the longevity of the instrument due to knocks and bumps, but it really would be taking things down the path of the resinator as Puff has suggested.

Cheers

Kim

[jeffhigh]

!/2" wide at the rims seems to be the norm
I presume you have seen this
http://www.dunwellguitar.com/DoubleTop/DoubleTop.htm
and this
http://www.reynoldsguitars.com/dtop.shtml

Randy Reynolds seems to use thinner plates and thicker nomex but then sands down the nomex
Both these guys are still using lower bout bracing

Seems like a lot of fussing around to me.
I do wonder whether it would be worth just laminating in a solid sheet of balsa?

[DarwinStrings]

The idea of Nomex appeals to me to a degree Kim and it could lead to a lighter soundboard, I am only just trying to get my head around the balsa lattice (even though I first listened to Greg about 17 years ago at Darwin Uni) but something new appeals to me so I might get around to Nomex one day. I did spend some time today taking a crash course in the material though cause you do have me thinking.

Do you know if the honeycomb luthiers using this method are using is phenolic foam filled or just straight honeycomb.

It is very light weight, about half that of balsa for low end balsa but around the same weight as extreme low end balsa. There is also a new Kevlar honeycomb which has better strength to weight than Nomex apparently, not that strength seems to be a issue here.

The balsa lattice thought was just a quick thought that I had not given any thought too but after looking at the weights (non filled Nomex) I found that you could, with balsa on its extreme end of its weight variation get the same weight as Nomex honeycomb (48Kg/m3) using end grain balsa (that is if you could use the same amount of glue) you could further reduce the weight using a longitudinal balsa lattice.

Jim

[Kim]

Yep, seen them both sites Jeff and the Donewell site is a good read, but the original fused kidney, 'braceless' in the lower bout design I referred to, was on a free Yahoo Geocities site. Yahoo started charging for domain names a few years back so it must have got sucked up in the vacuum of cyberspace with so many others because I cannot find it now. Like I said, I 'did' have images, but that was on the old XP system and when I done the upgrade to win7, I had transferred all my images to a .5T USB back up (never again USB backups cannot be trusted).

Months later when I went to access those images, I discovered that Windows7 had issues reading image files that had been transferred via USB using XP OS e.g. the file name gets changed AND so to does some of the load code in the image file that identifies it as a jpeg. The file still takes the full space on the drive, it will even still display the icon for the program you have selected to open .jpg files, but as soon as you click it gives an "unknown file type" error. I tried some recoveries cause I lost lots of images of my kids and the rest of the family and stuff but then the USB drive itself stopped working.....As a result I have grown to HATE microsoft, that episode was right up there with a house fire..so much lost my next will be a mac.

Anyhow you will just have to take my word for it. NO bracing past the mid transverse and as close to a true monocoque as I have seen, so it has been done. How successfully? No idea, but it was in one piece in the images. By the way, my 40mm rim specs where only a guesstimate from a long time back, the Donewell site I first seen probably 3 maybe 4 years ago.

Cheers

Kim

[Kim]

The balsa lattice thought was just a quick thought that I had not given any thought too but after looking at the weights (non filled Nomex) I found that you could, with balsa on its extreme end of its weight variation get the same weight as Nomex honeycomb (48Kg/m3) using end grain balsa (that is if you could use the same amount of glue) you could further reduce the weight using a longitudinal balsa lattice.

Jim

Good we got you think'in Jim

I think the issue you would have, even if you did manage to get just the right bit of balsa is that the glue surface would need to be sufficient to support the wood. Nomex on the other hand has a 'very' fine edge to the form. This would require very little glue and it is the glue which carries the most weight.

NOMEX_0358.jpg (25.35 KiB) Viewed 20610 times

Cheers

Kim

[DarwinStrings]

Yes I can see that and also that the end grain balsa might suck it up as well.....So that pic you showed is the stuff they use? no filling? it does have a very small glue edge, seems a bit scary to me though I suppose all up it is a reasonable amount. You still have not put me off the longitudinal balsa lattice idea though

Jim