Necks, truss rods, CF etc - Musings

[sebastiaan56]

In Localele's post on Camphor Ukes he mentioned a Cooktown Ironwood reinforcement for the neck. I believe that some early Gibson Mandos (A4's) used an ebony strip in the neck as reinforcement. A neck needs to be rigid, light weight and comfortable for the player. Seems to be a balance that could be easily over engineered. In "Left Brain Lutherie" David goes into some depth on measuring and controlling bend in necks.

Obviously neck length is a big factor, consider an Oud vs a Fender Strat, I cant imaging Hendrix's pyrotechnic on a 10fret instrument...... It depends on what you build.

So, the questions;

The fingerboard is probably the most stabilising factor for the neck given commonly used neck/fingerboard combinations. It seems to me that it could potentially perform most of the vertical / lateral stablilising. Why reinforcement?

Apart from ease of mass manufacture and the lower cost, why the steel truss rod? If a neck is rigid enough in its own right it shouldnt move, if its manufactured straight enough would it need adjusting.

A neck would tend to bow upwards from tension and twist from the uneven tension from the high and low strings. I believe the upward bow would be more pronounced but does bowing in two planes make a strong case for two truss rods? if any at all given a sufficiently rigid fingerboard?

Laminated necks seem the ideal solution to me. My friend the architect claims the T or I beam to be the ideal structure but would present problems for the player. A 5 piece with the two laminates either side of the centre laminate carrying the load and eliminating twist may be the ideal structure. This would remove the need for either CF or Truss Rod reinforcement.

Im sure many of you have been through this thought experiment before, where did you end up and why?

[Allen]

All this thinking is making my head hurt.

My Size 5 just has carbon fiber rods. The plans called for just one ebony reinforcement strip. Short neck and it's rock solid. Even now that you can get a truss rod to fit them, I don't think I'd bother the next time around. I don't know much about short scale instruments, so perhaps this is the norm in them.

As for a standard length neck in guitars, I haven't had near enough experience to offer much of an opinion, other than I've got some where the truss rod is most definitely an asset in controlling the amount of forward bow due to string pressure. These necks were ones that didn't get CF rods.

Now that all my necks are going to be an adjustable one, I'm committed to using 2 CF rods to help support the fret board from the 14 fret to the sound hole. Every person that has looked at my guitars that started asking construction questions has asked about the truss rod. I'm afraid that if you tried to sell a guitar without a truss rod, you have a very tough time of it. No matter how well made the neck was.

I still like the look of laminated necks. Mostly for the ascetics offered by using different colored woods and veneers. I really should spend more time thinking about how I stack those laminates together, but that starts to make my head hurt some more.

[jeffhigh]

Sebastian,
Whilst the commonly accepted opinion is that laminated necks are stronger/stiffer, from an engineering pov that is not necessarily the case, unless you are using stronger/stiffer woods in the centre laminate in widths that are sufficient to actually make a difference.(not just a 2mm stripe)

Laminates can provide resistance to twisting if you bookmatch them across the centreline. I build my uke necks this way in two pieces bookmatched on the centreline. No truss rod

Whilst we would like to consider the fretboard to be a major structural element it has a few things working against it.
- the material is subject to movement with changing humidity
-once the fret slots are cut a major part of the thickness is gone (the old Martins with bar frets were highly regarded for stable necks because the bar frets are in full contact with the slot unlike tang frets)

I built my OM without an adjustable rod, but with a CF rod epoxied into the neck towards the back of the neck and a cooktown ironwood fretboard. So far it is holding at 6 thou relief.

A Dead Straight neck is not normally the preference for a guitar, so building a hugely stiff neck that will stay absoluteley straight is not necessarily the answer.

I am by no means certain which way I will go on future builds, but I do feel that some form of extra renforcement is essential on a steel string neck, laminating won't do it by itself.

[Localele]

The Ironwood strip in my uke necks is 8mm wide and varys in depth from 7mm at the nut to 13mm at the dovetail body joint. It could be a bit deeper at the body but I am refining the necks and didn't want to run into it somewhere.Stiffness doesn't seem to be a problem with the comparatively low stress of Nylgut strings but I think I will be using truss rods in the guitars.Still have to decide which truss rod would be the best and would welcome any advice.Cheers.

[Allen]

The majority of people that have used them will tell you that the truss rods from Allied Luthiers are the best ones out there. I co-ordinated a group buy about a year ago now with a lot of the active members on the forum.

[graham mcdonald]

Lots of ideas there, but the really important thing to remember is that wood absorbs and loses moisture continuously as the weather changes, and different woods absorb and lose that water at different rates and so expand and contract differently. I have seen some guitar necks that need quite a radical truss-rod adjustment summer and winter, and others that never need touching. Given the option, I would rather have a guitar neck that has adjustable re-enforncement than one that doesn't, but Martin used nothing but a length of steel tube for many years and that worked fine. On a neck as short as a mandolin, I can't see the need for an adjustable rod, and a chunk of CF bar is more than enough.

String tension is more or less the same across the fretboard, whether a guitar or a mandolin set, so that isn't really a factor in warping necks.

I have an inherent distrust of laminated necks, most for the reasons mentioned about of different moisture absorbtion. I am uncomfortable with glue-laminating two or three different timbers which are all going to move at different rates. I see it as just building in problems for the future.

Sebastiaan mentioned old Gibson mandolins with an ebony strip laminated as a centre strip down the neck. There are an awful lot of those where the ebony has just disintegrated over the years (I suspect because it is a timber which is more hygroscopic than most) and it leads to a pretty complicated repair job.

Laminated necks are very much the fashion these days, which I suspect has to do more with the current materials fetish than any real advantage in building them that way. While cutting necks out of perfectly quarter sawn 4 x 4 timbers is mostly a thing of the past, a scarf jointed head and a stacked heel is a simple, elegant and stable way to make a neck.

I am open to debate on this

cheers

graham

[jeffhigh]

Sebastiaan mentioned old Gibson mandolins with an ebony strip laminated as a centre strip down the neck. There are an awful lot of those where the ebony has just disintegrated over the years (I suspect because it is a timber which is more hygroscopic than most) and it leads to a pretty complicated repair job

I have read that this was actually a dyed wood rather than ebony and that the dye is responsible for the deterioration.

I am with you graham on the muticontrasting neck laminations- not my cup of tea.
Dont mind the centre strip on my old Guild 12 string but beyond that is not for me.

[graham mcdonald]

Some of the old Gibson centre strips might well have been dyed something else, but I am pretty sure that I have seen a couple actually in half that looked more like ebony in the middle than anything else. First to admit identifying old bits of wood can be tricky

Ebony does get very brittle with age and it does keep shrinking. Over the years I have come across any number of old mandolin and banjo fingerboards noticeably narrower than the neck itself and crumble away when trying to remove them or just when getting the frets out.

cheers

graham

[sebastiaan56]

Thanks for the input guys, I'll keep on pondering,

Whilst the commonly accepted opinion is that laminated necks are stronger/stiffer, from an engineering pov that is not necessarily the case, unless you are using stronger/stiffer woods in the centre laminate in widths that are sufficient to actually make a difference.(not just a 2mm stripe)

Of course, any lamination would only be the sum of the parts, a 2mm strip would potentially be a liability as it greatly increases the surface to be glued, I would regard it as a decorative element. However, the advantage of using smaller amounts of precious timber is also worth considering.

Laminates can provide resistance to twisting if you bookmatch them across the centreline. I build my uke necks this way in two pieces bookmatched on the centreline. No truss rod

Makes good sense bookmatching, it deals with the environmental issues neatly as well.

Whilst we would like to consider the fretboard to be a major structural element it has a few things working against it.
- the material is subject to movement with changing humidity

Now this has got me thinking as all parts of the instrument would move by different amounts with humidity. This must affect the laminations in the neck as Graham points out. Are there tables for how Aussie timbers move with humidity? We should be able to work this out. The other factors would be the quality of the glue joint and the thickness of the sections of laminate. The bookmatching point is well taken as it would counter some of the opposing stresses.

-once the fret slots are cut a major part of the thickness is gone (the old Martins with bar frets were highly regarded for stable necks because the bar frets are in full contact with the slot unlike tang frets)

Yes, the thinnest portion would be the thickness at the bottom of the fret slot. You make a great case for bar frets this way.

A Dead Straight neck is not normally the preference for a guitar, so building a hugely stiff neck that will stay absoluteley straight is not necessarily the answer.

Please elaborate Jeff, my understanding is that under tension the fretboard should present close to a parallel surface to the strings. This may cause its own set of problems if you are playing with alternate tunings but the difference is probably negligible.

While cutting necks out of perfectly quarter sawn 4 x 4 timbers is mostly a thing of the past, a scarf jointed head and a stacked heel is a simple, elegant and stable way to make a neck.

I guess thats partially whats got me pondering Graham. Im in reductionist mode and trying to balance the amount of work (and potential for learnings) with the engineering required. Materials are a real issue indeed. Big slabs of quartersawn timber are the stuff of folklore.

[DarwinStrings]

I have been thinking about your post Sebastiaan and the answer was going to end up as a thesis, so can that idea, I'll hit a few brief points instead.

You are sort of right about the string being parallel to the fret board but you need to consider that when the string vibrates it does so i a log arch, just look at a vibrating string and you can see the curve in it. So what the fret board needs to do is be parallel to that arch. A well set guitar neck has a gentle curve and is not flat. I wish I had a dollar for every time someone brought me a steel string and asked me to fix the "bent neck" on it only to find a well set neck and and bad neck to body relationship instead, hence the high action which is why they want it fixed.

Necks twist both ways not just bass side up and it is usually to do with a crap piece of wood rather than tension, laminating can make a piece of wood more stable but if you use different species its a good idea to make sure the have a closely matched movement in service.

Books like "World Woods in Color" list "Movement in service" for all the wood in it including the Australia ones. Small movement is 0% - 3%, medium is 3% - 4.5% and "Large movement in service" well don't even consider that one unless you are building a new workshop. There is another one that is only Aussie stuff but I can't remember the title, I'll get it off my mate next time I see him and PM you the title.

Jim

It is a nice thought that a tree can sing when it's dead.

[jeffhigh]

Sebastian,

When evaluating a neck on a strung up guitar, put a capo gently at the first fret. hold the string to the fret at the heel(12th or 14th) and then look at the clearance from string to fret at the midpoint between these.
The string is acting as a straightedge and this clearance is called relief.
If there is no clearance you may even have "Backbow" which is very undesirable.

Most players will agree that some relief is desirable but disagree on how much
I like about 5 thou.

The string is never going to be parallel to the frets anyhow because clearance increases from minimal at the 1st fret to a few mm at the 12th

[Kim]

It is worth remembering that the neck of an instrument is more insulated against the effects of the variations in relative humidity or moisture in the air than any other part of an instrument. This is because the given volume of wood in a neck has the least amount of exposed surface area to the atmosphere. Also the area which is exposed is generally coated in a finish which, at the very least, will act to buffer against large swings in RH.

As far as laminations go, do not forget that the penetration of glue itself into the wood must be calculated in the equation when considering any added strength and stability that laminating timber will lend to a finished neck. More often than not the glue line and it's immediate surround will be MUCH stronger than the wood it is bonding together. Further more, in a laminated neck, much of the natural stresses that are found in a solid piece of timber have been released when the stock is ripped down to smaller dimension.

As the components are then bonded together to form a billet, the sum of their internal tension is much reduced. This is a basic principal of glue form lamination, they are indeed generally stronger and more stable than solid wood of the same dimension. If there is any negative to the process, and it is one worth considering in this application, it is the addition of weight to the finished product.

As for relief in the neck, it is a personal thing, however dead flat will produce fret buzz which some like SRV would accept for the speed, too much is just horrid to play.

Cheers

Kim

[DarwinStrings]

I thought I'd post the title of this other book here as it may be of interest to others. It has more Aussie woods that "World Woods in Color" and is more descriptive, a great reference, I'm going to do a quick search and see if I can buy a copy myself to save me borrowing it all the time.

"Wood in Australia" by Keith Bootle....I think i got his name right.

Jim

[Kim]

Yep, Bootle's book is a very good reference indeed.

http://www.dpi.nsw.gov.au/aboutus/news/ ... epublished

However like all other books that deal with wood desciption, it is focused at industry. The truth is that the variation of properties from one tree to the next within a species demands that it should really only be used as a general guide.

As an example, I have Tassi Blackwood as light as mahogany and mahogany as heavy as Walnut. I have Jarrah which is pale salmon pink, light weight and so soft that if you hit it on edge it will dent like pine. You can bet that stuff would move around with RH swings like the needle on a hair hygrometer.

Then again I have other Jarrah so dark chocolate it is near black and very very heavy. Hit that on edge and your back teeth will shake out of your freak'in head.

Cheers

Kim

[Paul B]

Bootle's book is a great reference, he's recently put out a second editition.

Not the cheapest book around, and now I have to buy the newest one.

[DarwinStrings]

Yes I agree there Paul I just searched and $90.00 will have to wait till after I get this X-mas thing behind me. Like many things I have been meaning to buy it for ages and it took this thread to get me off my rear end.

Wood can be frustrating like that can't it Kim, something should be done about it, can't these forestry people grow them under lights or something so they can provide us with a product that is the same as it is in the catalogue. What is wrong with them, I guess they are not trying hard enough, it's all about money and quality control has gone out the window.

Jim

[hilo_kawika]

If you don't want to shell out $90 but want a useful resource on woods and wood properties, you might try downloading the FREE "Wood Handbook" from the Forest Products Laboratory of the USDA:

http://www.fpl.fs.fed.us/documnts/fplgt ... gtr113.htm

Individual chapters can be downloaded rather than the entire book (14.3 Meg).

This is a thorough and relatively up-to-date publication (most recent references are ~ 15 years old) that includes information on some Aussie woods as well.

I believe that you will find in reading through this that a piece of laminated wood can be as strong and stiff as but not stronger nor stiffer than defect-free wood of the same dimensions and physical properties.

aloha,

David
www.ukuleles.com

[Kim]

I believe that you will find in reading through this that a piece of laminated wood can be as strong and stiff as but not stronger nor stiffer than defect-free wood of the same dimensions and physical properties.

Thanks for the link to the book David.

I had a quick read of what I believe would be the relevant text in your reference and it would appear that your statement is correct when one is assessing the strength and stiffness of jointed components in an assembly, but not so in the situation of beam laminations which is essentially the matter at hand in this topic.

Of course the term 'strength and stiffness' is relative to the material being bonded, the choice of glue, and the intended service environment of the structure but generally speaking, I find a side by side lamination to be much stronger and stiffer than defect free wood of the same physical properties. Further, if attention is paid to grain orientation, I believe a beam lamination to also be more stable.

Like most books which discuss the topic of wood, the "Wood Handbook" is understandably focuses upon the needs of industry. Considering the volume and variation of material it must cover and the many topics it needs to address, it is also understandable that the authors, as a matter of course, are forced to generalise to a certain extent. So like any of these books, I feel it should be used only as a general guide and once again, given the limited material that is used in the construction of an instrument, reiterate my point that each board should be assessed upon it's own merit rather than the specific description given to it's genus.

Strength and Durability

The ability of an adhesive to transfer load from one member
of an assembly to another and to maintain integrity of the
assembly under the expected conditions of service will
govern the choice of adhesive for a given application.

In building construction, adhesives that contribute strength
and stiffness during the life of the structure are considered
structural. They generally are stronger and stiffer than the
wood members. Structural bonds are critical because bond
failure could result in serious damage to the structure, even
loss of life. Examples of structural applications include
glued-laminated beams, prefabricated I-joists, and stressedskin
panels.

Bonded Joints
Edge-Grain Joints
Face-grain joints (wide surface of a board) are commonly seen
in structural laminated lumber products, where adhesive
bonds are stronger than the wood.

Cheers

Kim

[jeffhigh]

Kim
Laminated beams are exactly where the theory applies
In a stacked beam situation the glue is structural in that it transfers shear forces between the layers so that they act as one rather than just layers placed on top of each other.
In the engineering and building industry this enables you to use high grade timbers which would be impossible to dry in one large piece without cracking.
In the side by side lamination the strength is the sum of the components and yes you have the advantage of being able to orient grain direction.
You may also get a bit more stiffness from glue impregnation of the wood adjacent to the joint.
Think of it his way-
If I take a piece of timber 50mmx 50mm and cut it veritically along the length then glue the two piece together again, I would expect it's beam strength to be the same (except for the loss of the kerf width)
Now cut it in to 4 pieces and glue back together- same result
Where you can achieve extra strength/stiffness in a neck is if you use a hard and dense timber in the centre laminate with lighter timber to the sides.
Most of this centre laminate is cut away when the truss rod chanel is cut, but the layer left in the centre of the back of the neck is ideally placed to stiffen the neck without a hugh weight penalty.

[Allen]

We need a penny dropping emicon.

Thanks Jeff. When you describe it that simply, simple minds get it.

[Kim]

Ah yes kerrrpluk indeed Jeff

My mind seems still locked into spiral staircase handrails. About 25 6m x 90mm x 3mm strips all quickly glued and clamped around a 6m high timber drum form. One does not generally stop for smoko. The stiffness, strength and stability created in this type of application, (and yes as you say Jeff, it was indeed lateral loading against a vertical stack ) was quite incredible.

Just one point, the several hundred, knot filled, finger jointed bits of doug fir which make up the stacks in the two 9m x 305mm x 80mm glue lams currently holding up a large portion of the roof of my house could hardly be considered "high grade timber". I would think that more so than being used to avoid any drying and splitting issue associated with large solid timber beams, the main benefit of glue lam beams, aside from overcoming the scarcity of the former, would be the fact that the process adds so much strength that, not only can they be manufactured with just about any old reject scantling that once would have only seen the fires, but also when in stacked formation, they will also carry a greater load than solid timber of the same dimensions.

Cheers

Kim

[jeffhigh]

I used to be a structural engineer Kim, and designing with timber was something that I quite enjoyed, but I am way out of practice on doing any calcs now.
All structural timber is graded whether it be F5 for a piece of Radiata Stud timber or F16 for a good sound stick of hardwood and these ratings roughly eqate to the strength in Megapascals.
If you buy a piece of non structural Radiata, it has probably been rejected for structural use or just not put though the grading process.
Getting dry,defect free, unwarped and straight timber in large sizes and long lengths really is a problem and practically and economically unfeasible.
So the log is ripped down to strips which can be kiln dried, graded and faced and then reassembled into beams of any size as required.
An F8 beam is F8 beam for structural purposes whether it is solid or laminated from F8 graded strips.

[Localele]

While you are setting up for laminating necks don't forget to use up all those backsawn boards for the laminates.They end up quartersawn in the glueup.Also when glueing for a bolt on neck you can extend the centre laminate to be the tenon for your neck and glue up the neck block to become the mating mortise.

[DarwinStrings]

I have a question for Jeff based on the quote that Kim pasted from the online book that David posted.

The book states...... "In building construction, adhesives that contribute strength and stiffness during the life of the structure are considered
structural. They generally are stronger and stiffer than the wood members."

Some background information to set the question.......We are in a world where the tool makers are so good that they make a saw that cuts a kerf of 0.0mm.

If we take that piece of 50mm x 50mm and turn it into three pieces that are identical. then rip two of them up the centre and glue them back together, one using PVA, the second using Rescorcinol.

My understanding of the quote from this Book is that structural glues are generally stronger and stiffer than the wood. From that I assume that the Rescorcinol glued piece would be stronger and stiffer than the uncut piece and the PVA glued piece would gain no advantage.

From a Engineers perspective is one stronger or stiffer than the other, could you show a difference between the two even if you had to use a extremely sharp pencil?

Jim

[hilo_kawika]

Hi Toejam,

Actually there is a way of approaching the solving of your question by thinking about the approximate strength of the glue in the joint relative to the wood.

For arguments sake let's assume that the glue has the strength of a graphite carbon laminate. I believe that is a considerable overestimation but we're just creating a "straw man" at this point. On my website, with help from Josh Gordis, I made the beginnings of a spreadsheet to determine the effects of substituting different materials for the neck, fretboard and insert on how much the neck would bend under different string tensions and scale lengths:

http://www.ukuleles.com/Technology/neck.html

The spreadsheet is there to download.

So let's do the following calculation. First just see what the neck deflection is with a mahogany neck and an ebony fretboard and no insert. Next, use graphite carbon properties and make the insert depth the same as the neck (imitating the glue seam) but make the width 0.05" or about 1-2 mm (which would be a pretty healthy glue seam width) and compare the deflections from the two cases. Then double the glue seam width. I would guess that graphite carbon properties are probably 3-5 times greater than the structural glue properties but I have no hard data on that. In any case if the differences between the neck with and without the "glue seams" are very small then the effect of the glue seam on neck strength is negligible, right?

aloha,

Dave Hurd
www.ukuleles.com