Early F2 & F4 cracks

[Roccus]

I've noticed that a great many of the early F2s and F4s that I've seen in person and in photos have an identical top crack that runs from the base of the body scroll to near the edge of the inlaid sound hole rings. Does anyone know what it was about the structure of these mandolins that made this so common? Also, how stable are these cracks and do they require reinforcing?

[sunburst]

There is a cross-grain glue joint between the top and the head block inside the mandolin. Wood moves in response to changes in moisture content resulting from changes in relative humidity. Wood movement is not in the direction of the grain. In reality, there is very slight movement in the direction of the grain, but it is so slight that we can ignore it in general woodworking. Wood moves across the grain.
When we have a glue joint where the grain direction of the two pieces deviates very far from the same direction, there is always a danger of one or both of the pieces splitting when conditions are dry and the wood shrinks. The cross-grain glue joint restricts the movement of the wood, so it cannot shrink like it "wants to" and it can split because the tension that develops is beyond the tensile strength of the wood grain.
Old F2 and F4 mandolins have often been subjected to many cycles of humid and dry conditions, and many of them have been subjected to extremes of humid and dry conditions. The crack will usually show up right at the bottom of the curve under the scroll where strength is lowest because there is less wood to split.

In luthery, we make quite a few cross-grain glue joints. When trained wood workers are first exposed to the traditions of luthery, wood workers who have been taught to avoid cross-grain glue joints, to allow for wood movement in all joinery, many of them are shocked to see the glue joints we use. Instrument's are surprisingly durable considering how many things are "wrong" in their construction, and considering how casually they are treated by many owners who don't understand wood movement and glue joints. Owners who say things like "Lloyd Loar (or Stradivari, or Orville Gibson, or some other builder from history) didn't have any artificial climate control, and there are plenty of those instruments still around".
The fact is, if all of those old F2 and F4 mandolins had been kept in 40% to 50% relative humidity from the time they were glued up until now, many fewer of them would have that common crack.

Structurally, it can be a problem because there is a compressive force being applied to the top by the strings pulling the neck toward the tailpiece. That pull "tries" to sheer the wood along that crack. The head block is usually enough support to handle it, and many mandolins don't have structural failures because of that, but sometimes it becomes a problem.

[jim simpson]

I guess it would be awfully challenging to use an A-model neck block in place of the F-shaped neck block as the scroll is made from part of the neck block. It would be interesting to see a scroll made without the presence of the tradional F-block, just pondering.

[Tobin]

John Hamlett's explanation on humidity and wood expansion/contraction is likely the most influential factor in these cracks appearing over time. But wood movement alone likely wouldn't cause it if the instrument were just sitting around with no strings. It's the wood movement while under load that does it. There's a stress concentration that occurs along that line on either side of the sound hole.

The strings do indeed pull the neck up, putting the top plate in compression (and the back in tension) along the grain. But more likely, the locations of these cracks are due to what I would consider a structural design defect in oval-hole mandolins. And yes, I'm a registered professional engineer in structural engineering, so I can call it a defect with a straight face. :)

The cracks usually occur exactly in line with the edge of the sound hole (I've seen them on both sides, not just the scroll side) because that's the area where shear stresses are concentrated from the force of the neck. Unlike later mandolins with an elevated fingerboard, the old oval-hole mandolin extensions were connected to the top almost to the edge of the sound hole. So as John mentioned, when you try to pull the neck up from the nut end (as the strings are doing when under tension), it will want to rotate around the fulcrum at the neck block. Which means everything on the tail-side of the neck block is rotating downward. Due to the location of the sound hole, there's nothing to offer resistance except the shear strength of the wood in its weak direction (wood is always stronger in the direction of the grain, not transverse to the grain).

So the shear stresses are concentrated in this area, right from the widest part of the sound hole up to where the top meets the neck block and side supports.

If they had given the oval hole mandolins an elevated fingerboard extension to keep this rotational force from affecting the top, these cracks would likely not occur, even with normal wood movement from humidity changes.

There's also a component of force coming from the bridge area, due to the location of the sound hole and the brace, forcing all the down-force from the bridge to transfer around the sound hole, which is another contributing factor to the locations of these cracks. Basically, the sound hole is put in the worst possible place for carrying normal loads on the top.

[Roccus]

Thanks. Great answers, guys.

[Tobin]

Just adding more thoughts to consider...

The F5 design, with its ff-holes on either side of the bridge, and its elevated fingerboard, eliminated this structural flaw in the previous Gibson style mandolin top. It took the rotational neck force off the top, keeping it constrained to the neck block, and it gave the down-force from the bridge a continuous load path along the grain. The sound holes are not in the way any longer.

Hence the fact that F5s have "tone bars" and not "braces". Notice that F5 bars are running almost perpendicular to the bracing of an oval hole. The F5 bars run parallel to the grain, offering virtually nothing in the way of longitudinal strength. The wood grain and arching of the top takes care of that already, and we don't see the same cracking points on old F5s. Aside from being an acoustical engineer, Lloyd Loar had enough of an engineering mind to realize that the oval hole design was poorly thought out for structural integrity.

[sunburst]

We do see similar cracks in F5-style mandolins, as well as the "point crack" that originates at the corner of the bottom body point. I believe both of these common cracks have less to do with forces resulting from string tension than from "poor" wood working techniques. The point crack results from a stress riser set up by the sharp, inside corner on the end grain part of the top whereas the scroll crack results from the cross-grain glue joint, and the mechanism for causing the failure of the wood is, I believe, wood movement.
The oval hole may contribute to the formation of the scroll crack, and the f-hole may contribute to the point crack, but we see both cracks in both styles. String tension may contribute to both, but I don't think it is the main cause of either.

[Roccus]

Does this make sense? The crack starts at the top over the scroll block due to the varying grain direction expansion rates. Then, once started, progresses toward the sound hole as a result of the string pressure on the neck and top.

[Bob A]

Regarding reinforcing: I've had my 1921 F4 since the early 1970s. It has all three of the cracks mentioned; the treble side crack at the soundhole seems to have been glued, probably well before I got it. The bass side has the appearance of a crack, but it does not go through the top, and is stable. The lower point crack is over an interior block, and does not move. So my experience with them is that they are a non-issue. If you can induce any movement at the crack, get it glued; otherwise, no worries.

The only attentions the instrument has needed from a luthier in 40+ years were getting the top brace reglued - it had become partially detached - and two episodes of the back coming loose from the ribs around the tailpiece area. The latter was due to the gradual shrinking of the back, and was addressed by removing a section of binding, adding a narrow sliver of maple to the edge of the back to size it to the rim, and regluing the binding. A bit of dye to match, and a spot of French polishing, and you can't even see it unless you know where to look.

[Tobin]

We do see similar cracks in F5-style mandolins, as well as the "point crack" that originates at the corner of the bottom body point. I believe both of these common cracks have less to do with forces resulting from string tension than from "poor" wood working techniques. The point crack results from a stress riser set up by the sharp, inside corner on the end grain part of the top whereas the scroll crack results from the cross-grain glue joint, and the mechanism for causing the failure of the wood is, I believe, wood movement.
The oval hole may contribute to the formation of the scroll crack, and the f-hole may contribute to the point crack, but we see both cracks in both styles. String tension may contribute to both, but I don't think it is the main cause of either.
Yeah, I agree. I think the across-the-grain expansion/contraction over the blocks is what's causing these cracks to start. My 1918 F4 and 1917 H4 both have cracks in these exact areas, but they are small and stable. I would think that typically when they start to appear, there is an immediate stress release, and they will stabilize. Gluing them should generally keep them from spreading. But on mandolins where they have spread, I'd guess it's probably from that stress release causing the internal forces to shift around from string tension, leading to what I mentioned earlier.

If you do see these same cracks appearing in the same places on old F5s (which would make sense, as the internal blocks are the same), do you see them spreading down towards the middle of the top plate like you do on oval-holes?

[sunburst]

Sometimes, I think the transverse brace coming loose inside the oval hole mandolins "helps" the scroll crack lengthen. It's sort or a chicken and egg situation, though. Did the crack help the brace come loose or did the loose brace help the crack lengthen... or a little of both?
I don't recall seeing a scroll crack extending very far into the top of an F-5 style mandolin, probably, as you say, because of the lack of a good destination like a big oval hole right in the string path.

FWIW, we see cracks beside the fingerboard extension in guitar tops too. They usually crack right at the edge of the head block, and if left unattended will continue to crack into the sound hole, and then will start to sheer (like the San Andreas fault) from string tension. It's a slightly different situation, though, because the head block doesn't bridge across the crack like it does in mandolins.