Where did the metal go?

[evfreek]

Hi. I was trying to join two pieces of 1/2 inch round with 3/8 inch rivets. First, I punched two holes in the main piece spaced 3 and 3/4 inches apart, then a single hole in the secondary piece. The holes were perfectly centered, but one was a little ragged . I figured that I would mark the last (critical alignment) hole cold, then drift to align.

Unfortunately, this reflected poor planning, because it does not account for the growth of the bar when punching :confused: . It does account for thermal expansion, since both measurements are done cold. If the last hole is punched at 3 and 3/4 inch spacing, the final spacing will be larger due to the metal flow from hot punching.

Since I was not able to do a test bar, the next best thing is play-dough. I made a bar of the same dimension, and "punched" a hole with a dowel, leaving the slug in the hole. As expected, the metal around the hole swelled out on the sides of the bar, identically to the hot punched steel. But the hole did not move as much as I thought it should. The side swell was about 1/16 inch, and the lateral hole movement was less than 1/32 inch, more like 1/64 inch. Measurements this small are difficult to perform on clay.

The next best thing was an area based calculation. The area of a 3/8 inch hole is 0.11 sq in. If the bar is approximated as a 5/8 inch diameter circle and it is assumed that the thickness does not change on punching, the "bar" will grow radially by 0.1102/(pi x 5/8) = 0.056. Thus, the 1/16 inch observation is pretty close for the sides.

So, a bar is different than a circle . Where does the extra metal go? Is it somehow restrained by the absence of a free surface (boundary) near the hole? Is it growing into the third dimension (thicker)? This is hard to measure with a ruler, since it does not take much swell to eat up this volume.

Anyway, this exercise shows that for two holes, at least, it is probably safe to ignore the slop. Shortening the measurement by 1/64 or less seems safe. This amount of slop is pretty straightforward to drift out anyway, and a hot rivet will fill it easily. It does make one curious, though.

[Chris Pook]

Are you punching the holes or slitting/drifting? Punching as in knocking out a slug, or slitting/drifting with no material loss?

[evfreek]

Hi Chris. I am doing punching O not slitting []. There is a slug. It is less than 1/16" thick. It is very clean and disk shaped and is slightly dished. This accounts for a small amount of the missing material, but very little. If you do the play dough trick all the way to the bottom (can't do with steel ) the results do not change much.

[evfreek]

Hi. I think I found out what is going on. It is what I initially suggested :cool:

To further bolster the reasoning, I carried out two more experiments with play dough. First, a 5/8 inch disk was punched. This showed the nearly 1/16 inch increase in radius (on all sides). Then, a short 5/8 inch bar was punched. It showed an approximately 1/32 inch growth in length. This is consistent with the increase in spacing of 1/64 inch.

Apparently, I do not have a good feeling for even a gross estimate of the results of a simple linear plastic deformation in 3 dimensions. Not to worry, nearly nobody else does. Back to steel, I punched the ornamental piece with a 1/64 inch shrink, and the holes lined up just fine .

[jj2k]

There ya go, asking questions and then answering yourself! Sounds psycho-analytical to me..... (grin)

[mcraigl]

Why not punch both holes on both bars at 3 3/4". Seems that if they are the same stock dimensions they should shrink/swell/expand the same amount if you're using the same punch? If you mark 'em both cold and punch 'em both hot seems like it should work.

[evfreek]

You are exactly correct, mcraigl. That is what I should have done. I realized this after I did the deed. It is easy to see how it happened. Before the bars were punched, they were flattened, and were a bit too hot to lay out. Should get a bigger pair of blacksmith calipers. The error was realized after the deed was done.

No real harm done, and the experiment was fun and instructive. It was very surprising to me that the amount is so small, since from a volumetric basis, it seems that it should be larger.

Is it normal for two bars riveted together to have some wiggle? I am afraid that this will be a vulnerable crack for corrosion in the future (outdoor application). I have not done any of this larger ornamental garden work before.

[HWooldridge]

I have seen an adjustable trammel setup where you punch the first hole in a bar, then place a properly sized dummy in that hole (one end of the trammel) the other end is a scribe to give you the proper line across the bar, which is then punched with a conventional center punch. The scribe is held in a small block that will slide along the trammel and lock in the proper place. Similar to blacksmith's dividers but a bit more precise.

If done hot with proper top and bottom tools, rivets will hold very tightly because they shrink and pull the pieces together. This assumes the pieces are relatively cool (room temp or only slightly hotter) and the rivet is somewhere in the orange to low yellow range. This is one of the reasons they are still used in some structural work and were once the standard for boilers and other plate structure, like ships.

If I want a rivet tight, I typically set up the whole thing cold and heat the rivet with an O/A torch, then set it. The mating pieces seldom get so hot as to compromise the fit. If you want the rivet to move - as in tongs - you put it all together then heat the assembly. Once riveted, the tongs can be worked for a good fit as they cool.

[Pault17]

Thanks HW! I hadn't made that connection (no pun intended), but after reading it, it makes many much more sense. Until I learn to forge weld, or any kind of weld for that matter, mechanical connections are my main method of joining.

[evfreek]

Hi Hollis. Thanks for the very helpful tip. This sounds like exactly what I should do. Time to get out the torch! But it is an oxy-propane, not oxy-acetylene. One of the rivets has not pulled quite tight on the joint, and that is exactly where the looseness is.

I have a lot of work to do on my riveting technique, since until now, it has only been for tongs.