Oud Action Adjustment with a Compression Rod?

Quote: Originally posted by jdowning

In a recent discussion about guitar truss rods and their function (chaldo forum topic 'Screw in the pegbox into the fingerboard what purpose') it was supposed that such a device would have no application for ouds - a short oud neck being relatively stiff and resistant to bending. An oud or guitar neck is essentially a structural beam subject to mostly compressive and bending loads. These loadings - due to string tension - apply an offset (unbalanced) compressive load to the neck joint (about equal to the total string tension) as well as an offset bending force (bending moment) that tends to bend the neck upwards at the nut end. The simplest of the guitar truss rods is a compression rod - a screwed rod that when tensioned applies an offset compression load to the neck - in opposition to the string forces. By adjusting the compression rod the neck of a guitar can be made to overcome the string forces causing a neck to bend backwards (deflect) and so reduce string action over the fingerboard. For a uniform simple beam subject to pure bending loads the maximum deflection at the centre of the beam is proportional to the bending moment and length squared (i.e. length X length) and inversely proportional to the material stiffness (measured by Young's Modulus) and the geometry of the beam cross section (Moment of Inertia). In other words - all else being equal, a longer guitar neck will deflect much more proportionally than a short oud neck when subject to bending forces. The question is how much will an oud neck deflect under an offset compression load and - if so - is the deflection sufficient to be of any use in adjusting string action? Or does a compression truss rod have any useful application in an oud? Wood being extremely variable in its physical properties, it is futile to attempt engineering calculations based upon the average published properties for wood species. The only way to gain an understanding of the possibilities and limits is to carry out some testing on a model of an oud neck. More to follow!

Hi jdowning :-)

In some guitars (with the much longer neck as an Oud has) sometimes the main piece of wood the neck is made of.... is laminated by seperate tiles of the wood itself. Means, the stiffness / rigidity is maximized by each glue-joint between the single tiles, according to make the neck stiffer as it would be if made of one part of wood. The question ist, does it affect the tone-quality of an oud when implating such a trussrod ? The weight of the rod might be less if using a slim one, but does a such rigid part in the neck make the pegbox-neck-joint beeing critically ? One has to adjust perhaps, and has to reach a srew or something adequate, so the neck has to be a little more thick at the end, and so at the neck-body-joint. This, if the soundholes are "closed" by a rosette... if not, you may reach a screw from the inside of the bowl in the neck-block. It might be worth a try, doesn´t it ? Therefore the neck has to be constructed as a complete mechanical device... like as it´s done on a guitar-neck, if glued or bolted.
I saw some examples of an Oud with adjustable neck... in two different styles... but I think, the neck has to be mounted the same way as always, in other case the adjustment only has effect on the action... but what´s with the rest of the sting(s)... behind the finger or over the top of an Oud ?
In normal a guitar neck is going a specific lenght into the body, and the the action is normaly adjusted at the 12th fret... but the fretboard is going farer... on an Oud the adjustable lenght will be only the length of the neck... so you have to "bend" only the neck. This may cause sometimes buzzing, when you make the action lower... although the action at the neck-body-joint is correct...

Another question is, if the maker has to fix the neck (fingerboard) with an angle to the soundboard or not... when using a "trussrod"...

Hundreds of ideas...

Fritz
It might be, that anyone has to create a Oud-able trussrod

Thank you for your thoughts Fritz

My objective here is more straightforward - that is to investigate if a simple compression rod set in the neck of an oud might be designed to produce sufficient deflection of the neck to be of any practical use in adjustment of the string action. With this in mind I have only two questions to answer at present.

Question #1 - can a compression rod apply sufficient force to bend the neck of an oud? The answer is yes but not very much.
To test this idea I first made a simplified experimental model of the core of an oud neck from pine - semicircular cross section, length 20.5 cm, width at neck joint 50 mm and 40 mm at the nut. A slot was milled underneath parallel to the upper surface to accept a standard 1/4 x 20 screwed rod. The slot - with rod in position - was then filled with a pine strip glued in place.
The slot was made deep enough to provide an offset of the bolt centre-line below the neutral axis of the neck cross section.
The neutral axis of a semicircular section is positioned from the top surface a distance given by 0.42R where R is the radius of the cross section. The neutral axis is the plane where tension and compression forces - due to bending - balance. It is the plane around which bending occurs.
The offset of the rod means that when the rod is tightened the compression force induces a bending moment causing the neck to bow upwards in the centre - the greater the compression force applied by the rod and the greater the offset the greater the deflection. This force is limited either by the yield stress of the rod (i.e. the load at which the rod starts to irreversibly stretch (in the case of a 1/4 20 screwed rod this load is 520 Kg) or - more likely - when the wood cells of the neck fail under local compression forces. For the purposes of this test, in order to minimise the effect of localised crushing of the wood cells and maximise the observed bending of the neck under load, large washers were placed under the nut to distribute pressure.
The nut was then tightened by hand to a tension that felt about maximum load (about one turn of the nut).
The test neck upper surface was observed to have bowed - a total negative deflection at the nut end measured against a reference surface was 0.75 mm. Not much! (This deflection would be less for a stiffer wood or more for a less stiff wood)

Next - Question#2 What will be the total deflection at various compression loadings with the neck core (fitted with a fingerboard) glued and bolted to a neck block?

Quote: Originally posted by Fritz

It might be, that anyone has to create a Oud-able trussrod

I don't know if a mandolin compression truss rod could be made to work on an oud. Note, however, that the commercially available mandolin truss rods appear to be designed for use on the Gibson style flat bodied mandolins with solid guitar like necks and not the old lute like Neapolitan mandolins with their separate neck and neck block.

For these trials a screwed rod has been used for simplicity and because this set up can probably exert more bending force on a neck than a mandolin style can. Even with this arrangement and with a softwood neck core the total deflection in bending is only about 0.75 mm. With a stiffer solid hardwood neck the deflection in bending would be proportionally less.

The next stage has been to test a new configuration this time with the compression rod passing through the neck and neck block - an arrangement that changes the loading so that not only is the neck subject to bending loads (as with a conventional compression truss rod) but the neck joint is subject to compression loading.

The attached images show the new test rig. The softwood neck core has been fitted with a boxwood 'fingerboard' - 2 mm thick and glued in place. A boxwood plate has been glued at the 'nut' end - to simulate the hardwood of a pegbox. The softwood 'neck block' has a thin plate of rosewood glued to the 'neck joint' surface to simulate additional hardness/stiffness due to the ends of the ribs in a real situation. A piece of spruce has been glued to the top of the block to simulate sound board end grain that will add to stiffness of the neck joint.
The 'neck' was then glued to the neck block and the top surfaces planed level. In order to measure total deflection under compression load a flat particle board plate was screwed to the 'neck block' surface - deflection being measured with engineers 'feeler' gauges at a distance of 18.5 cm from the neck joint.

The screwed rod at the inside face of the neck block was fitted with a double lock nut and large diameter washer to prevent crushing of the wood under load. For simplicity and purposes of these trials the lock nut was anchored in the jaws of a vice while the rod was loaded by rotating the nut at the other end with a wrench.

Results to follow

The attached image is a plot of the compression rod load against total deflection measured at a point 18.5 cm from the neck joint over three loading cycles.
As I have no means of measuring the compression load, the loading is represented by the fractional turns of the compression nut - up to a limit of 1 full turn. Total deflection measured is in millimeters.

Test #1 - curve AB is a straight line returning to the initial zero point deflection of 0.35 mm after the compression load is released. Maximum deflection is 3.4 mm. The straight line indicates that the material of the neck/neck block assembly is elastic under these loading conditions.

Test #2 - curve ABCD. The loading initially again produced a straight line (AB) from a zero deflection of 0.35 mm but the maximum loading was sustained for 24 hours after which time the maximum deflection had increased from 3.4 mm to 3.7 mm. On releasing the load the zero deflection had increased to 0.63 mm an indication that there had been some degree of permanent yield of the wood cells under maximum loading. In other words - under these time related loading conditions - the test assembly is semi-elastic.

Test #3 curve DC is again a straight line with loading applied to full load and then released without any time delay.

The bend in the neck alone under full compression load measured 0.45 mm but the bending also results in a slight rotation of the neck at the neck joint face causing an increase in the measured deflection relative to the neck block. The remainder of the deflection under compression loading - occurs at the neck joint due to local semi elastic compression of the wood cells. The compression loading is eccentric, due to the compression rod offset, concentrated in the lower part of the neck joint - hence the downward deflection of the neck.
(If the rod had been installed along the neutral axis - i.e. without any offset - then there would be no neck deflection caused by the compression loading condition)
Note that in order to produce a maximum deflection of about 3.7 mm at the pegbox end of the neck would only require an equivalent closing of the bottom of the neck joint - due to cell compression and/or the neck joint face rotation - of only about 0.25 mm.

String tension will apply a counteracting bending moment that would tend to lessen the deflection somewhat - not accounted for in these trials. However, the effect of string tension (say 40Kg load) is likely to be much less than that of the maximum load that might be applied by a compression rod.

The purpose of these tests has only been to investigate if a simple compression rod - if run through both neck and neck block - might produce significant deflection of an oud neck to have some practical application as a device for correcting high action due to 'pull up' of a neck under string loading over time.
The amount of deflection that can be achieved for a particular oud design application will depend upon a number of factors including the wood stiffness of the neck and neck block and the relative compression stress characteristics of the wood cells.

The compression rod used for these tests - for convenience - was a piece of 'off the shelf' 1/4 20 screwed rod purchased from a local hardware store. This standard rod has screw pitch of 20 threads per inch so that one rotation of a nut will move it about 1.25 mm. A better solution - allowing finer load adjustment - would be to use a plain rod screwed at one with a finer pitch thread.
Also, in a practical application, the other end of the rod must be welded or otherwise securely fixed (to prevent rotation) to a pressure bearing flat plate which is then anchored to the inside face of the neck block with screws.

One further observation.
These trials have shown that a compression rod produces a relatively small deflection of the neck (due to bending of the neck) in comparison to deflection of the entire neck due to compression of the wood at the neck joint.
It should be noted that for these trials a piece of rosewood veneer was 'sandwiched' - cross grain - at the neck joint. Wood is much weaker in compression perpendicular to the grain than it is parallel to the grain (about 10 times weaker for some species). It is likely, therefore, that compression of the veneer insert itself may have been responsible for much of the neck deflection observed under the loading of the compression rod.

This suggests that insertion of a thicker (5 mm say) cross grained piece of wood of selected species at the neck joint might allow a more controlled downward deflection of the entire neck at a much lower compression loading. This would in turn mean use of a smaller diameter compression rod.
This action adjustment, of course, only provides a lowering of the string height (which is what is usually required). Should a raising of the action be necessary then it would simply be a matter of replacing the insert with one slightly wedge shaped as appropriate.

Further examination of the geometry of the test piece indicates that for a deflection of 1mm at the nut end of the neck the movement at the bottom of the insert under compression loading would only have to be about 0.1 mm assuming that the rotation at the neck joint is about the neutral axis - or for a deflection of 3 mm, about 0.3 mm. Measurement of any compression of the insert between the unloaded and loaded condition was attempted using small pair of draughtsmen dividers, a magnifying glass and a good quality steel ruler graduated in 0.5 mm increments but the results were inconclusive - a movement of perhaps about 0.2 mm being estimated which at least might tend to confirm that the insert was indeed being compressed (according to plan).

At full compression load there was no measurable bending of the neck longitudinally.

The attached plot of three cycles of compression load against neck deflection from zero load to maximum load (1 turn to 1 1/4 turn of the compression rod nut) show that the insert has been compressed beyond its elastic limit for each cycle. In other words when the maximum load is released the insert does not return fully to its original dimension and the neck takes up a permanent deflection. So - in the case of the first loading cycle - the neck takes up a permanent deflection of about 0.2 mm when the compression load is released. For the second loading cycle (starting at a permanent deflection of 0.2 mm) the permanent deflection on release of the compression load becomes about 0.4 mm.
For the third loading cycle (starting at a permanent deflection of 0.4 mm) the compression load was increased to a maximum of 1.25 turns of the nut. On release of the load a permanent neck deflection of about 1.25 mm was recorded.

Note that in each case time was allowed for deflections to stabilise (24 hours). So, for example, in the case of loading cycle #1 the deflection at full load was first measured as 1 mm. Measured 24 hours later this deflection had increased to 1.25 mm indicating a plastic deformation of the insert.

This test confirms that a simple compression rod (with compression insert) may be used to control neck deflection in order to lower action at the neck joint. At low compression loadings the deflection may be fully reversible. At higher compression loading the neck deflection may be partly permanent, partly reversible.
On the other hand a neck reset, each time, is 'once off' permanent - non adjustable - whereas the flexibility of adjustment with a compression rod may have a useful application in avoiding a need for costly repeated neck resets?

Here endeth the tests!