Official Luthiers Forum!

"The comment I made earlier about dowels through the bridge had me thinking about what you are talking about in the above post for most of the day while working in the shop. For the sake of conversation what about a mortise/tenon between the bridge and bridge plate (matching materials and grain). Sounds like a cool CNC project/experiment build. Can anyone think of a reason not to try it? It won't prove anything but with a sinker WRC top, in theory should give peace of mind if successful. On the other hand it presents a ton of problems/challenges with flush glue joints and very tight tolerances unless the tenon were to come from the bridge and extend through the bridge plate."

A simpler solution might be small light weight bolts through the tie block area of the bridge, top, and bridge plate but this is a solution eschewed by most luthiers. I have done that for pinless bridges where I wanted to keep a small footprint.


"Since it's the stress at the edges, and in particular the back edge, of the guitar bridge that fail and cause the thing to come up, the longer the distance between the front and back edges of the bridge the longer it's likely to stay put. From this standpoint it doesn't matter whether the belly is in the front or in the back, and I suppose there's something to be said either way. I just think Martin got it right the first time, and Gibson was simply trying to be different. "

Having more wood in front of the saddle may help reduce the risk of the forces acting on the saddle from breaking out the front of the bridge. But I do think Gibson bridges are a bit uglier.

It's easy to minimize the tipping force on the saddle; just angle it back. If the saddle bisects the break angle there's no net tipping force. This also increases the downbearing on a UST a little. Rick Turner says that a nine degree back angle automatically adjusts the compensation length if you change the action height by raising or lowering it, too. I've used a nine degree back angle for several years now, and have no complaints since I got calibrated on figuring out where to put the bridge.

I'm coming to this thread late, but for those that would like a picture, here are some FEA plots showing the stress distributions on the glue surface:
http://www.mimf.com/phpbb/viewtopic.php ... &start=100
So, the glue line stresses for the torqued bridge are indeed similar to the pure shear case described above by Alan.
A main point of that FEA model was to find how the ball-end stresses are propagated through the parts. But IIRC, I did a few trials with a shorter bridge from front-to-tail, and as expected the peak stresses were higher.
On page 8 is a more complete model -- there is a slight stress riser at the 'corner' of the belly, so the Gibson bridge is probably slightly better, but I didn't do that case and I have even less an idea of Gibson's motivation.
Someday I'll get back to that model and explore more variations.

_________________
David Malicky

Thanks for those models: I'd forgotten that. Did you take the friction between the string and the bridge as the string turns down into the hole into account in any way in the model?

Good question. In that model, I assumed the string maintained the same tension all the way to the ball end. My thinking was that over time, with sufficient vibrations, the tension would equalize throughout the string. That's typically (but not always) how mechanical systems behave with friction and vibration. If that's not true, it would be fairly simple to apply smaller forces at the ramp and ball-end, and friction shear forces at the ramp and saddle.

Of course, then we'd need to know how the tension is shed at the friction points -- if you have any data on that, I'm glad to hear it! I'd guess that the bass strings would be most likely to shed tension, simply because the rougher surface could hang up on the saddle. A model could be developed to simulate it, but it would be difficult to accurately account for the effect of vibration and especially the wound strings. A tiny load cell between the ball-end and bridge plate could measure that force, but the extra mass of the load cell would affect vibration. Maybe calibrated 'crush-washers' would be light enough, or a tiny custom strain-gauged transducer.

_________________
David Malicky

The load cell would at least settle the question of what the forces on the ball end actually are. I've often thought of putting a little piezo element in there to see if I could actually measure a signal at that point. It would have to be done on a rigid beam, of course, to eliminate any accelerometer effects...

Alan, that would be nifty!

New thread on the interface stresses: viewtopic.php?f=10101&t=46993

_________________
David Malicky

Thanks for making that comment Alan, I was wondering the same. As discussed in the pinless bridge thread I'd be curious to know if a model for a pinless bridge would have any affect on the results, especially in the case of WRC vs a typical spruce.

Good question. My guess is the stress distribution changes for WRC would be subtle, and the natural variation in shear strength is probably greater than the average difference between WRC and Spruce. IIRC there are also good orthotropic material properties for WRC and those could be studied in the FEA model.

_________________
David Malicky