neck heel crushing sides of guitar

[Brian Evans]

I built this guitar with a bolt on neck, a square tenon, maple neck, mahogany sides, and I kind of relieved the inside of the neck heel so I could finesse the neck angle a bit easier. Over time the neck heel is visibly crushing the softer mahogany, and changing the neck set angle. I fixed it once with shims, the shims have crushed out. Has this happened to anyone else? Oh, I am using strings in the 160lb tension range.

[Aaron Helt]

How tight are you torquing the bolts?

[Barry Daniels]

How about light gauge strings?

[Aaron Helt]

String tension pulls the heel away from the body.

[Alan Carruth]

It sounds to me as though you're over tightening the bolts. This can not only crush the sides but also slit the heel, depending on what sort of fittings you used and how they're oriented.

Several years ago I made a 'test mule' Classical guitar with a demountable neck, so that I could use it again on a different one if I wanted to, and for airline transport. It's held on by string tension, with a single bolt through the heel to a T-nut in the neck block to hold it from tipping up. There's a gap between the bottom of the heel and the side, and if I don't put in a wedge it's easy to inadvertently tip the neck back while playing and bottom out the strings on the fretboard. I can push the hell in against the body with one finger. Arch tops use more overstand, so there is not as much leverage at the heel, which would require more force, but not much more than twice as much. So, two finger's worth of pressure? How tight do you need to torque the bolt to get that?

On my bolt-ons I use furniture bolts with big heads and Allen wrench sockets. I just make them finger tight, and they're fine so long as everything fits.

[Brian Evans]

Keeping in mind that as an archtop, the T-wrench that I use to tighten the allen big head furniture bolts is 25" long and therefore twists madly when I put torque on it, and I am using thumb and a couple of fingers on the tee handle to twist it, the bolts qualify as finger tight. Two bolts, the lower bolt holds the heel in, and that lower part of the heel doesn't move. I suspect that when I take the strings off so I can put my wrench in and check the upper bolt will be loose - it was last time I did that when I put the shims in. The wood of the sides that the neck rests against is visibly crushed in. From my calculations there is about 5 - 7 lbs of "lift" at the heel, and the 135 - 165 lbs of tension pressure from the strings forcing in. I do normally use a set of strings that produces around 135 lbs - .012 - .056, tuned down to D-D. I think I will calculate the tension pressure from a few inch-lbs of torque on the 1/4" NC thread and actually see what that bolt torque is. This all just seems very odd to me.

How interesting. I knew that bolts can generate a lot of axial clamp load, but actually looking into it is kind of fun. For a 1/4" bolt, steel on steel, unlubricated, the ratio is 1 inch-pound of torque produces 20 lbs of axial load. Now I need to measure or calculate the torque I am generating, which is problematic with a 25" long spring between me and the bolt head. I think I can calculate the spring rate of the steel rod.

Brian

[Alan Carruth]

I remember reading an article years ago in the Exprimental Aircraft Association mag about calibrating C clamps. The guy wanted to be sure e got even pressure all along a 30' spar, so he made up a fitting for his torque wrench to grip the clamp head and then checked each one using a bathroom scale. He wrote the conversion factor for each one on the clamp. The day of the glue-up he got all of the members of the local chapter in with torque wrenches, and they got the whole thing clamped up and properly torqued within the working time of the glue. Maybe you could use a similar tactic?

[Brian Evans]

Measuring with a torque wrench, best guess is I am finger tight at 20 inch pounds, maybe 25. That is light (5 lbs, or one finger pressing lightly) pressure on the end of a 4" allen key. That yields around 400 lbs of initial clamp load, which seems like a lot. But even .010" crushing or settling of the wood in the joint will reduce that to effectively zero, leaving only the added pressure of string tension (which is what Belleville washers are for). Which explains why when I de-string the guitar and check the bolt tightness, the upper bolt is fairly loose. You tend to think that the neck tension is trying to pivot around the top of the side, lifting the bottom of the heel away from the body, and sometimes it is. But when you adequately constrain the heel from lifting it becomes the fulcrum, and the neck tries to push at the very highest point, hence I see the crushing of the sides. Also, the change happens over quite a long time. In this case, around 5 months before the change became noticeable.

[Pat Foster]

Can you post some pics?

Pat

[Mario Proulx]

Are you certain that it is the wood being crushed, and not the neck block(and ribs) rotating?

[Brian Evans]

When I get a chance I will take it apart and take some pictures, but you can visually see the dent in the wood of the sides. With the neck off it's very apparent. I strongly suspect that my mistake was to hollow out the cheeks of the neck heel where it bears on the sides so that I could floss the neck into alignment easier. I bet the cheeks of the neck heel are no more than 1/8" wide where they press on the sides. My real surprise is that no one else has experienced this - I usually make common mistakes, not unique ones!

Brian

[Mark Swanson]

I seriously doubt that is what is affecting your action. The sides would need to be seriously crushed enough to make a difference in the neck angle, I bet it is something along the lines of Mario's question. I have seen many necks cut like yours, and they have no troubles.

[Mario Proulx]

I've done a couple hundred necks that way, with zero issues....

Highly suspect you have a rotating neck block...

[Barry Daniels]

What kind of mahogany is this? Philippine is very soft stuff and could crush in this situation.

[Brian Evans]

I really have no idea what kind of mahogany it is, I thought it was Honduran. I don't know how to tell different types of mahogany apart. I took the neck off and took pictures. I can measure the indentation at between around .020" at the deepest point, which is enough to change action height noticeably. You can see pretty clearly the dent in the lines in the binding, and one side is worse than the other. You can also kind of see the shim material that I stuck in almost disappearing near the top. I can confirm that I did the bolts up tight, maybe around 40 - 50 inch pounds, which is far more that string tension, and probably around 800 lbs of preload per bolt. What I can also confirm is that on the side that is having the worst issue the heel cheeks have been relieved so that the surface pressing on the side is very narrow, almost angled all the way out with no flat bearing surface at all. I think those two problems are what caused this, and I can fix them both. I may leave the top bolt out, and just do up the bottom bolt lightly after I fix the sides and put it back together.

[Barry Daniels]

The relieved edges of the heel should be flat where they bear against the sides, not a knife edge.

And it sounds like you are over-torquing the bolts. I would take them to finger tight (where the heel makes contact) and then go about 1/4 to 1/2 turn more.

[Pat Foster]

What Barry said, plus, from what I can see in your second and last pictures, the cheeks of the neck aren't making full contact with the sides like they should. It looks like most of the force from the bolts is concentrated at the top, near the soundboard, and at the bottom, near the heel. Along with the knife edges, the force needs to be spread up and down the sides evenly from the soundboard to the back. If one side is depressed more than the other, it could mean there's a high corner, like a table with one leg too long. The cheeks should sit squarely and steadily against the sides without gaps, even with the bolts loose.

Looks like most of the denting is happening near the soundboard. If it's from overtightening the bolts there, keep in mind that it's the bottom bolt, near the heel, that has to do most of the work in resisting string forces that want to pull the neck up.

[Beate Ritzert]

What would be the repair?

Trying to cut the flanks of the mortise more even, those of the tenon as well, shim those and fit the neck a 2nd time? Or is there a simpler solution?

[Brian Evans]

It's clear that I made several obscure but fundamental mistakes with this neck joint - fair enough, it was my first one. I'm basically going reset it, clean up the cheeks, add some material to fill in where I undercut the cheeks too much, stabilize the side wood, and put it back together with more care on the bolt torque, probably leave out the top bolt entirely. One issue I discovered this morning is that part of the bottom of the tenon (near the back) is actually bottoming out on the floor of the mortise, keeping the cheeks from bearing on the sides near the bottom of the joint. There is lots to fix, ample opportunities for learning from my mistakes...

[Brian Evans]

All fixed. Reset the neck, filled in the undercut cheeks with some epoxy putty, cut the mortise back so that the tenon isn't bottoming any longer, and boom like that the joint became as stable as you could ask for. Just used the one bottom bolt. Sounds great with new Newtone low tension bronze wound strings, too.