Which: re-rivet or weld?

It sounds like there are too few rivets, each carrying too much stress. Is there room for more rivets?
Airplanes are built with a combination of high-temp-cure epoxy and rivets, precisely to spread the stresses out.
Welding around the existing rivets won't fix an overstress problem.
Is the boat made of a weldable alloy?
Joe Gwinn

This is a tricky one. That's not really how aircraft rivet-bonding works (the rivets are there only to prevent lifting of the edges, and resulting failure in cleavage and peel), although that *is* how rivet-bonding of high-strength steel works, in modern automobile assembly and repair. We can discuss this if you're interested.
It would work nicely on a small boat, but it would be pretty tricky, because epoxy bonding to aluminum is extremely weak unless the surface is properly prepared. It also would depend on having plenty of overlap at the joint.
If it were me, I'd double-up on the rivets, then drill out the old ones and replace them, too.
I had such a Sears jon boat, back in the '60s. Mine was built by Grumman. Sears used all kinds of suppliers, even including Fairchild Aircraft.
FWIW, I think they all are. Most are made of 5052 or 5083/5086. Some 6061 is used for ribs, etc. All are weldable; the 5XXX series is more common and more weldable.

Or it's a really old boat.
I was told -- and I don't know if this is true -- that truck frames are riveted because that allows for a little bit of flexing on overload, so the frame may get tweaked but it won't get ripped apart.
I don't know if this applies to the boat or not, but my inclination would be to either keep it all riveted, or make it all welded, but don't muck around with welding up some rivets and leaving the rest in their virgin state.

I have an old Jon boat, from Sears, that has a bunch of leaky rivets. In the past I had one of these boats, not a Sears but almost identical that also had leaky rivets. I re-set all of the leaky rivets and after a few years some started to leak again. No surprise there. So with the boat I have now I'm thinking of MIG welding around all the rivets on the outside of the boat and then running some beads on the inside bracing because some of the rivets will be loose. I know the welds will add some drag but I really don't care, the only bodies of water this boat will be going in are small so even if the motor quits the rowing distance will be short. Opinions? Thanks, Eric ===============================================================
I would be really afraid that when you weld on the head of a tight rivet, the heat cycling will allow it to loosen, and what you will wind up with is all loose rivets with the heads welded to the outside sheet. The rivet is under tension now, heat it and it will weaken and if it gets too weak it will yield to relieve that tension, and when it cools it will be loose - that's my theory, anyway. I agree that you should double up the number of rivets, add the new ones first then drill out the old ones so you wind up with twice as many all new rivets.
----- Regards, Carl Ijames

I did know the roles of glue and rivets in airplane manufacture. (I have been reading Aviation Week for since I was 13 - my father subscribed.)
I didn't know about the use for steel in automobiles. I am interested, actually.
I wonder how the Russians built the MIG-25 fighters, which were made of steel foil. Search. Wikipedia says that they used a weldable nickel steel alloy.
Yes to all of it. I wasn't suggesting this approach here, but was making the point about the necessity of stress spreading. For one thing, one would need to drill all rivets and disassemble the boat, to allow adequate surface prep. Apply goop and rivet back together, then put in the autoclave and bake at 300F overnight. No autoclave? Hmm. That is a problem. The airplane companies have all manner of clamp-and-heat fixtures for this.
So we would end up with three times the number of rivets? That ought to do it.
My recollection was the same, that 5xxx series alloys were used, which would make it weldable. But my sight-unseen guess would be that one would do seam welds of the sheets to one another, and ignore the rivets. This actually sounds like a good application of gas welding, and was used to make airplanes in WW2. Though they used billions of rivets too.
Joe Gwinn

=============== Why not just clean it up good, goober the seams and rivets with silicone seal (inside and out if desired) and repeat as necessary. This ain't a space capsule.

Sandpaper and JB weld.

Jerry Mayfield, who was an editor at AvWeek in the late '70s and early '80s, and I used to compete to scoop each other on aerospace manufacturing techniques (I was two floors below him, at American Machinist). He had the advantage of being weekly, to my monthy, but I beat him on single-crystal turbine blades and bismuth-alloy brazing .
Try this for a starter:

They're trying to avoid working and loosening the rivets -- quite different from the objective in aircraft.
Good question. They couldn't use titanium because they hadn't figured out how to seam-weld it (we used electron-beam welding in a vacuum, starting, I think, with the F-111).
I don't know how they wound up joining it.
Yeah, but you can get over half the strength of A-B cure high-temp epoxies with good room-temperature-cure material. That should be plenty.
Hmmm. I meant double the number. I guess that was ambiguous.
It would be good for gas welding -- for someone who is really good at it. I consider myself lucky to weld two aluminum tabs together; without punching holes. But Kent White (the Tinman) has some great info on it.

this was my solution on a very old canoe, worked but i always brought silicone along for long trips

I believe that the SR-71 came first :-)
>I don't know how they wound up joining it. > >> >> >>> It would work nicely on a small boat, but it would be pretty tricky, >>> because epoxy bonding to aluminum is extremely weak unless the surface >>> is properly prepared. It also would depend on having plenty of overlap >>> at the joint. >> >>Yes to all of it. I wasn't suggesting this approach here, but was >>making the point about the necessity of stress spreading. For one >>thing, one would need to drill all rivets and disassemble the boat, to >>allow adequate surface prep. Apply goop and rivet back together, then >>put in the autoclave and bake at 300F overnight. No autoclave? Hmm. >>That is a problem. The airplane companies have all manner of >>clamp-and-heat fixtures for this. > >Yeah, but you can get over half the strength of A-B cure high-temp >epoxies with good room-temperature-cure material. That should be >plenty. > > >> >> >>> If it were me, I'd double-up on the rivets, then drill out the old >>> ones and replace them, too. >> >>So we would end up with three times the number of rivets? That ought >>to do it. > >Hmmm. I meant double the number. I guess that was ambiguous. > >> >> >>> I had such a Sears jon boat, back in the '60s. Mine was built by >>> Grumman. Sears used all kinds of suppliers, even including Fairchild >>> Aircraft. >>> >>> > >>> >Welding around the existing rivets won't fix an overstress problem. >>> > >>> >Is the boat made of a weldable alloy? >>> >>> FWIW, I think they all are. Most are made of 5052 or 5083/5086. Some >>> 6061 is used for ribs, etc. All are weldable; the 5XXX series is more >>> common and more weldable. >> >>My recollection was the same, that 5xxx series alloys were used, which >>would make it weldable. But my sight-unseen guess would be that one >>would do seam welds of the sheets to one another, and ignore the >>rivets. This actually sounds like a good application of gas welding, >>and was used to make airplanes in WW2. Though they used billions of >>rivets too. >> >>Joe Gwinn > >It would be good for gas welding -- for someone who is really good at >it. I consider myself lucky to weld two aluminum tabs together; >without punching holes. But Kent White (the Tinman) has some great >info on it.

I had a really old aluminum canoe that had been retired from rental service in Algonquin park that had been bashed around, taken over beaver dams, and generally well abused before I got it which had 2 or 3 gashes welded and a few missing rivets "plug welded". In the 15 or more years I subsequently owned it, it never leaked or squeaked. I sold it to a friend who used it at their family cottage for another 8 years or so, and sold it with the cottage 2 years ago. This was a pretty heavy "springbock" canoe. Not sure how it would work on a lightweight Grumman, but I have no reason to believe there would be any problem, with the Grumman being 6061 about half the thickness.

Putting sealant on the rivets does not work long term. I have tried this in the past. I have tried several different sealants and the only one that worked for more than a year was this black stuff that remained gooey forever. Eric

Thanks everyone for the suggestions and opinions. Yesterday I decided to try welding over a couple rivets because I was welding some cracks anyway. MIG welding didn't work because of crud trapped under the rivet head and TIG welding did OK because I was able to boil some of the dirt out but was still a bad solution. Since there over 100 rivets I'm going to order a couple hundred closed end blind rivets and borrow an air powered rivet puller. I'll drill out the old rivets, clean up the area some, dip the rivets in sealant made for the purpose, and install 'em. I found a few sealants that are made for use under a rivet but none that work outside for any length of time. JB Weld will not work by the way. I know all about abrading aluminum surfaces wetted with epoxy, and it does work, but this method can only be used on the outside of the hull on this boat. And epoxy patches on the outside don't fare so well with beaching the boat. Eric

Not just for airplanes. I've seen aluminum brazing used to fabricate aluminum heat sinks for water-cooled radar transmitter modules.
Thanks. I've printed the article out for reading.
Or the SR-71, as has been pointed out.
The Russians do have the capacity to forge an entire fighter airframe from titanium, which eliminated the need to weld titanium.
In the US, I recall plans to have an entire factory with an argon atmosphere (and people in spacesuits) to work titanium. Don't know what happened though.
Titanium is pretty springy, with lots of spring-back. Don't know how well titanium rivets work.
I did a lot of gluing for NASA as a summer employee in the late 1960s, and we used 180 F (82 C). Higher is better, but many electronic components could not tolerate soaking at 300 F for very long.
Copy editor sought. Apply within.
I bet welding is also pretty slow compared to riveting.
Joe Gwinn

Right. furnace brazing is, or was, quite common in aluminum, and automobile radiators have been some of the biggest users.
The bismuth-alloy technique is something different. It applies to joining superalloys, such as Hastelloy, and the P&W job I watched it being used on was for turbine-blase halves. It may work with other alloys; I don't know.
The brazing foil is made of Hastelloy with a small amount of bismuth added. The bismuth lowers the melting point; the objective in this case is to drop it about 50 deg. F.
The parts and foil are placed in an oven at something like 10 deg. F below the melting point of the parent metal. The foil metls and the bismuth diffuses into the parent metal. As it does, the melting point of the foil rises; the temperature is raised a few degrees; and the joint solidifies at that temperature.
The bismuth has no significant effect on the melting point of the parent metal. The finished joint has a melting temperature within around 5 deg. F of the parent metal. The finished assembly is virtually a solid piece with uniform metallurgy.

My brother and I went on a tour of the SR71 at the Boeing air and space museum in Seattle. The man giving the tour was an ex SR71 pilot and he had some great stories. One of the interesting things he told us was that when building the first ones titanium was hard to get in the USA. At least titanium of good enough quality to build an airplane. So the CIA, who was having the plane built, bought titanium from the Russians. They had to go through some pretty convoluted channels to buy the titanium without the Russians knowing where it was headed for and who was buying it. The pilot also told us how when flying over the Soviet Union missles were fired at them and they watched on radar as the missles tried to catch up and hit the plane and then fall back to earth without ever hitting even one SR71. A great tour, it lasted over an hour, and there were only about 15 people on the tour so it was real personal. Unfortunately the tour didn't include us actually being able to sit in the plane. But it was still one of the best tours I have ever experienced. We heard lots of really interesting stuff, the tour guide was a good speaker and a good raconteur. Eric

Ahh. I see from the following that you weren't talking about this kind of brazing. I'll have to ask the mechanical guys what brazing alloys were being used.
This is almost a kind of welding. I've seen the rough equivalent done in the making of silver jewelry, where a series of silver brazing allows with melting temperatures about 100 F apart are used so the workpiece can be built up in stages. Also, it's well known that it requires a higher temperature to un-braze a joint than the original braze job, and jewelers depend on this difference.
I've read it. Very interesting.
Don't know how widely used in production this is, but the effect on fatigue life is dramatic - from 10^4 cycles to 10^8 cycles. Although not given is the effect of varying the number and arrangement of spotwelds. Arrangement in particular is very important, and a circle of welds is far better than a line.
Joe Gwinn

I think you're going to see some interesting developments in car manufacturing over the next couple of years. Advanced high-strength steels (AHSS) are taking over and welding them is difficult. Thus, rivet-bonding and weld-bonding are being promoted as solutions.
Based on a preliminary look, it appears that the manufacturers are also looking at laser welding for the new grades of steel. Some are using it now and my guess is that they'll take over. At the same time, direct-diode lasers are coming onto the market for welding and other jobs.
My Associate Editor wrote a piece this month on direct-diode lasers:

Tomorrow or Tuesday, our new magazine for lasers in manufacturing, _Shop Floor Lasers_, will go live. The editor wrote an excellent piece for the first issue on hybrid welding -- MIG plus laser -- and it's amazing how deep but narrow those welds can be.
I don't have a URL for the magazine yet but here's the place-holder homepage, which should have a link for the first issue by tomorrow or Tuesday:

If it's so old that the rivets are leaking you have more problems than just leaky rivets.
Like metal fatigue - concentrated around the leaky rivets.
I would not even bother trying to weld or rivet.
Maybe coat the bottom with truck bed liner?
Or recycle it for aluminum scrap?