It may not be obvious in the photo below, but this Ninety year old Gibson K-2 mandocello has seen its share of hard times. It's a classic relic from the days of the great mandolin orchestras, and it has the potential of doing its job once again, after some serious restorative effort:

Shown in strong reflective light, the top damage is quite evident:

A bit "under built" from the get-go, the top started to sag under the load of downward pressure at the bridge, particularly on the treble side. Even with the string tension released, the sagging are is very pronounced, to say the least:

Here's yet another view of the top distortion:

As you can see, the bridge has had a very thick shim added to raise the action as the top began to fail"

And, adding insult to injury, there's been some seriously bad "repair" done to loose areas where glue failure occurred. Notice the gobs of white glue above and the deformation and cracking of the sides.

Two more photos of the misaligned end block area. Along with the sagging area of the top, the instrument was close to complete structural failure at the tailpiece.

As if that weren't enough, there's an equally messy misalignment and glue job at the neck block area. This poor thing was really potted in white glue:

One thing for certain was the need to take the back off the instrument to fully assess the damage and put things back in order. Working around the sides with my favorite sharpened putty knife, I was able to pry the back loose, a little section at a time:

Most of the Gibson mandolin family instruments have the top and back binding right at the glue joint edge, so it isn't necessary to remove the binding to take the back off:

From mandolin to mandola to mandocello, these old Gibsons were built with only one small brace inside, just below the oval sound hole. As one might expect, the tops of the old mandolins and mandolas also fail occasionally, but a sagging top is more likely with the larger expanse of unsupported soundboard of the mandocello.

This brace was seriously loose, of course, contributing to the weakness of the top:

As expected, there was a nice mess of glue around the neck block:

Just for a bit of added "convenience" the glue had run way down between the sides and the curved neck block sections:

All that would have to be cleaned out, of course.

Day 2: The Plaster Mold

Before attacking the glue and other side damage from previous repair efforts, I decided to make the plaster mold of the top contour so I could set it aside to dry while I worked on other parts of the job. I'd be needing a good mold of the top contour to be able to recreate the original shape where the top had collapsed, and a plaster mold is the quickest and easiest way to get that mold made.

To protect the top finish, I covered the entire area with tin foil:

Now, it's important to know that I'm not using "tin foil" as a generic, meaning "aluminum foil." This is REAL tin foil, and it has a characteristic that's really desirable for this job. Unlike aluminum foil, tin foil has no "temper" and it's really limp. If you've never handled the stuff, you're in for a treat - it's so flexible that it feels almost like cloth. Unlike aluminum foil, I can smooth it down to fit the precise contours of a surface and it doesn't tend to spring back at all.

In case you're wondering where to get tin foil, I'll say that the only source of supply I've found is through dental supply companies. My foil is 1 mil thick, and I wouldn't want it thicker. My pal, Joe, has some that's 0.7 mils thick, and it seems even a bit nicer than mine. No, it's not cheap. When I called my local dental supply, they had no problem selling foil to a civilian. . .

 

Because I was about to lay a great mess of plaster of Paris on top, I made sure to seal the joint with some thin tape - I wouldn't want any moisture leaking through:

I mixed the plaster to peanut butter consistency, and formed a dam around the edge:

Then with some at mayonnaise thickness, I filled in the rest, mixing small batches so I could handle it comfortably before it began to set:

Once the entire area was covered, I waited about an hour for it to set up so it could be handled:

Then, taking that big hunk of plaster off the instrument, I turned it over, and glued it down to a 3/4" plywood work board:

It would spend the next week or so drying out fully. By the way,one patch looks gray in the photo above because one of my batches was from a different brand of plaster, so it came out darker looking.

Day 3: Sticky Old Glue!

The sound hole brace popped of with the merest jab of my putty knife, so I put it aside while I worked on the damaged sides:

Well, that white glue had run way down into the space between the loose sides and neck block, so I had to break out my steaming rig and squirt hot steam as far into the joint as I could to soften the glue:

Nasty stuff - but once it's softened it can be teased out like strings of freshly chewed bubble gum. It took a right smart while to get it all out of there, but with reapplications of steam and increasingly thinner blade palette knives, I got the job done:

Down at the butt end of this beast, the sides were much more damaged and the white glue much more liberally applied. I had to take everything apart here, so I used a hot knife directly on the thick glue:

Not a pretty sight!

Everything had been reglued out of alignment, and there was white glue everywhere, including the joint between the top and the sides.

Turning back to the neck block area, I spent some time making up special contoured clamping blocks to press the sides back into contact with the neck block. Those are the white maple ones with cork padding to avoid damage to the finish. Those blocks would slide right off the curved surface when clamped, so I made up a big block I could use to hold them in place. Once I gave my setup a quick dry run, I added some nice new hide glue to the joint:

Working quickly, I was able (with the help of an extra pair of hands) to clamp everything in a matter of seconds, before the glue began to gel:

Day 4: End Block

Regluing the end block was comparatively straightforward. After cleaning off all the old glue, I realigned the sides, slathered on some new glue and clamped it up, using a heavy wood clamping call with a number of saw kerf slots to allow it to bend to conform to the rounded contour of the end block:

To avoid gluing the loose top to the block or linings, I slipped a piece of thin steel shim stock in the loose area. I'd want to be able to do that regluing operation separately, when I'd be able to see alignment clearly.

Day 5: Regluing the Top

Some sections in the area near the end block had the top reglued out of alignment with the sides, and nearby portions of the same joint were loose, with the sides distended outward. First I broke loose all the offending joint area with my trusty sharpened putty knife:

Then, after cleaning out the old dried white glue, it was time to reglue the end block and sides to the top. It took a couple of big bar clamps to press the sides inward to mate up with the perimeter of the top, and more than a half dozen shorter clamps to press everything together after I pasted in some nice new hide glue:

OK, time to put it aside to allow the glue to dry. I usually leave it clamped at least 24 hours, so I'm certain that the glue is fully dried and solid.

Day 6: Patching a Hole in the Side

I decided to get some of the small stuff out of the way before attacking the collapsed arch in the top. The end pin hole and tailpiece mounting screw holes were enlarged and generally pretty messed up, so I thought it would be a good idea to fill them completely and redrill appropriate holes when it was time to remount the tailpiece.

A machinist's taper pin reamer in the ol' Makita made a good job of rounding up the end pin hole. I just ran it all the way through to get a good cylindrical bore:

A hardwood dowel fit nicely, and I glued it right in place - it took a few light hammer blows to seat it tightly:

Because the dowel was a tight fit, I could trim it with my chisel:

And I did pretty much the same thing with the other holes:

That nasty irregular hole would be another issue. I chopped the edges just a bit with a chisel to make them into a more regular shape, and then traced the outline on a piece of transparent tape:

With the tape on a thin piece of maple, I had a nice clear cut line to guide me at the band saw to rough out the plug:

After truing up the edges with a knife, I had a decent fit, and I glued it in place:

Unlike the dowels, this one needed to dry overnight.

Day 7: End Wedge

A short day today. I leveled the maple plug by slicing with my sharp flat chisel:

Then, I made up a new rosewood end piece, and glued it in place:

Naturally, it would have to dry overnight as well.

Day 8: End Wedge

An even shorter day - just a matter of trimming the rosewood end wedge with the flat chisel:

Tomorrow (or the next day) is a big day - time to work with the plaster cast and start to restore the top arch.

Day 9: Reforming the Arch

Take a look at the two photos below. I pressed the contour gauge down against the undamaged (bass) side of the plaster mold, and when I compared the arch to the treble side (right) I could see lots of daylight at the ends because the mold had a pronounced hump where it reflected the collapsed portion of the top

I figured it would be handy to have a contoured sanding block to help correct the treble side of the mold, so I laid some coarse sandpaper down on the bass side, and used that contour to form my sanding block:

My favorite material for this use is a high density polyurethane foam - it sands really easily and it's rigid enough to hold its shape. Cork works well, but is just a bit softer and more flexible.

Here's another pair of shots - one showing the fit of the sanding block on the bass side, and the other revealing that nasty hump in the mold on the treble:

Then, it was a matter of sanding and scraping to even out the contour:

The sandpaper loaded like crazy with plaster, so I did a fair bit of bulk removal with curved scrapers, constantly feeling the surface with the flat of my hand to keep from digging in too deeply in spots.

The sanding block made it easy to "fair" out the curve:

OK, now for the real deal I spritzed the inside of the top with water, to get it about as wet as I could without dunking it in the sink:

About an hour earlier, I'd stuck a sandbag into my old soup cooker, and covered the entire rig with insulating towels:

The idea was to heat the sandbag to around 150F - you know, "soup temp."

Then in went the hot sandbag - right on top of the wet surface. I squashed the sand around and pressed it down to distribute pressure as evenly as possible

For extra pressure, I added the rest of my sandbag collection:

Then, it was back upstairs for the top to settle in, cool and dry out. I left it under the sandbags for several days.

Day 10: Arch Restored!

Well, after some time under pressure, the top looked almost as though it had never collapsed. While it has regained its original shape, it certainly wouldn't be strong enough to carry the load of eight heavy steel strings tuned to pitch. A bit of reinforcement under the bridge could help hold things in shape.

Here's the original sound hole brace - it retained its original curvature, so I glued it back in place:

Day 11: Bridge Brace

Today's job was to make and glue in that brace to reinforce the top in the bridge are. It was a matter of hand-fitting the brace to the arch of the top, and then gluing and clamping it, using the plaster mold to support the top arch:

Day 12: Preparing to Reglue the Back

The back had come off cleanly, as those things go, but there was still some leftover glue to clean off the mating surfaces before it could be reglued. I spent some time with sanding blocks and scrapers to prepare the surfaces for regluing:

Hide glue doesn't adhere well to surfaces contaminated by dirt or oxidation, so I scraped the back as well, paying special attention to the areas where the glue had failed before:

A few bits of lining were either loose or missing, so I took care to reglue and replace them:

Day 13: Preparing to Reglue the Back

I wish I could have shown how much the sides were distended and out of alignment from the back, but I wasn't able to line up the shot adequately. Suffice it to say, there would be no way for me to be certain to have been able to reglue the back in one go, so I chose to reglue the neck block area first to maintain the best alignment I could in that area. If you look at the end block area you'll see I have a clamp there - it's to keep the center line of the back in agreement with the center of the end block and end pin hole. I made up a heavy padded caul to distribute pressure and clamped the back to the end block with hide glue:

Now with that heavy area tied down, I'd be able keep things aligned as I began to force the sides to line up with the back.

Day 14: Clamping & Regluing the Back

Despite the fact that there's only a single photo here, I must say that this particular day was the "big one." In order to align the sides with the back, I used every one of my right-angle bench clamps, pressing in like crazy all around the periphery. I also used a small bar clamp between all the bench clamps to press the back down and provide clamping pressure for gluing the back to the sides, linings and end block:

Just setting all this up took a long time, but because I had previously glued the back to neck block in good alignment, all the parts fit back together as I slowly increased inward pressure all around the body.

Day 15: Color Day

Because there was a good deal of tension in the clamping job, I chose to let it remain clamped for an extra couple of days before releasing the clamps. Once the clamps were off, I used hot water to dissolve and remove the excess glue:

After everything was dry, I went over to the stain cabinet, picked out some transparent dye (in MEK) and mixed up a batch of red with a little brown to approximate the original coloring on the maple back and sides.

Using a nice artist brush, I applied the stain to the raw wood where all the old stain had been damaged by water and previous work, and right over the old finish as well in some spots:

Then, I blended the color with a rag and some more stain diluted in alcohol:

As you can see, the stain went everywhere, even all over the binding. The stain dried immediately, and I wiped a few thin coats of shellac over all of it to seal the stain so it wouldn't be damaged by handling as I worked the next day.

Day 16: Scraping the Binding:

Now, for the last bit of color work - removing the stain from the binding. My favorite binding scraper is a small piece of hard steel with a 45-degree bevel on each end. Both ends are the same, and both are sharpened the same. I use a small diamond hone to put a really nice sharp edge on the end of the scraper:

I hold the scraper nearly vertical as I guide it with my fingers to scrape the finish and a tiny bit of the surface of the ivoroid binding:

This operation really cleans up the look of the finish and makes the colors really come alive!

Day 17: Finishing

This is one of those "days" that's really a series, with days of work, separated by several days for finish to dry. The finish is shellac, applied in the classic rubbing technique known for centuries as "French polishing." It's this final clear coat that levels out the finish and gives the whole instrument a fine glow.

Before adding finish, I scuff sanded the entire surface to allow the new shellac some extra surface area to get a real "bite" as it dissolves into the old finish. My first coat was very wet indeed, to make sure I got serious adhesion:

In the days that followed I slowly built a nice coating to achieve a good smooth look:

Day 18: The End

Some new strings and a bit of setup work were all that remained, so today is the day for some final "studio" photographs of the result - a fine looking and solid old instrument, ready to make some fine music:

There's no sign of that top failure now: