How long is a piece of string... or a rivet rather.

[Lisa]

Just a quick question; to make a round head on a ⅛" rivet how much length should be sticking out? I seem to recall it being something like 1-1.5 times the diameter, but can't recall exactly.

[baldric]

See here www.sapphireproducts.co.uk/641t1.htm, if you go from here www.sapphireproducts.co.uk/solid.htm you can find details of many head types

[Roger]

I don't have a lot of experience in this, but I think the standard information assumes that you want to create a full sized head for maximum strength. The longer the piece you have to form, the more likely it is to bend over as you form the head. If you aim for a smaller head, life gets a lot easier and the head can still look really good. It's debatable whether you ever need the full strength of that full head, so I'm not convinced it's worth aiming for.
If you can form the head with a press or in the vice, there's a lot less chance of leaving marks. The anvils can be Silver Steel in the soft state. I just used a ball nosed cutter in the lathe to make them. I only resorted to a press tool because there was no way to use a hammer and punch. I'll be doing all the rest that way too.

[suctionhose]

Can only be finalised by experiment.
Variables are hole size and material thickness. That is, how much length is needed to fill the hole before the head swells.

While 3/4 rivets would have a 13/16 hole in the small sizes it's best to make them neat - like a dowel - in the hole.

Cutting rivets is easily done with a home made shear. A couple of holes moving past one another. Mine consists of two bits of ground stock with a bolt through like scissors. Nothing fancy. Or you can hold them in a split bush in the lathe and part them or hacksaw them off.

While side cutters are fine (I've done thousands that way) for forming a visible head, you want a nice flat ended shank to start off with.

[RGR 60130]

1.5d = 3/16" in this case and see what happens! Adjust as required.

The traction engine boys with lots of wheel building experience can probably enlighten you further.

Reg

[baldric]

Sept 15, 2016 12:55:34 GMT suctionhose said:

Can only be finalised by experiment.
Variables are hole size and material thickness. That is, how much length is needed to fill the hole before the head swells.

While 3/4 rivets would have a 13/16 hole in the small sizes it's best to make them neat - like a dowel - in the hole.

Cutting rivets is easily done with a home made shear. A couple of holes moving past one another. Mine consists of two bits of ground stock with a bolt through like scissors. Nothing fancy. Or you can hold them in a split bush in the lathe and part them or hacksaw them off.

While side cutters are fine (I've done thousands that way) for forming a visible head, you want a nice flat ended shank to start off with.

It I worth pointing out that for 3/4" rivets the hole has plenty of clearance as this will be hot riveted I expect, thus the rivet will be bigger than the 3/4".
I would use the tables I linked to as a guide, as has been said there are several variables that can come in to play such as hole that you have to fill as the rivet swells.
Baldric

[SteveW]

If I recall its a simple volume thing. You need enough sticking out to provide the volume necessary to fill the rivet head you want.

Key thing if you've got a number of rivet heads to form is to make sure they are all the same size. Maybe practice technique and tune length needed.

[Jim]

Sept 15, 2016 15:27:02 GMT RGR 60130 said:

1.5d = 3/16" in this case and see what happens! Adjust as required.

The traction engine boys with lots of wheel building experience can probably enlighten you further.

Reg

Building the Burrell I used the protruding length of 1.5 the diameter to get a nicely formed rivet head. As others have said a nice close fitting hole is important too for the rivet body to 'swell' into so creating a nice strong joint.  For the very small rivets I use a set of nail punches to form the heads as they have nice concave ends intended to locate on the nail heads and they are cheap too...always a plus.

Jim.

[Lisa]

I had a bit of an experiment, tried various lengths, and ultimately a bit under 1.5 the diameter worked best, in terms of easy to form and a fairly decent result; which is what the tables baldric linked said. The actual job meanwhile is half-done while I make a suitable snap for getting into the tight corners.

[suctionhose]

I thought I'd resurrect this thread to post the latest on rivet lengths from my workshop...

I'm working on a pair of traction engines, ploughing engines that is (with an accomplice). We're making the rear wheels at the moment. A pair of engines has 4 rear wheels obviously with a total of 1152 rivets (3/16 steel)!

Thankfully 860 odd have been counter sunk for the strakes so no special head forming required.

For the spoke ends though, a round head, otherwise known as a Snap head, has to be formed. Theses heads are the visible ones so a neat job is essential.

My net plate thickness is 0.430" and I find I need 2 x dia (0.375) of shank protruding to form the head to the proper size. Why more than 1.5d and where does the metal go?

Well, after looking at the volume of a truncated spherical heads being 0.75d high x 1.75d dia in accordance with some age old, presumably British Standard, I find that 1.5d of shank is equivalent to the volume of the finished head with nothing extra to spare. Great in theory but nothing is allowed for the rivet to expand in the hole or deformation of the joined plates. In other words, the guide of 1.5d is highly theoretical and most unlikely to be the actual allowance required.

Given that I need 30% more metal in the protruding shank than the 1.5d rule would provide, I started reading Poisson's Ratio and Hooke's Law to see if Steel was at all compressible in the state of plastic deformation. I think not. Seems the volume is unlikely to change so one must assume the metal "goes somewhere".

Answer to "where" is swelling of the shank and dimensional variation of the finished head. Even with the rivet a tap-in-fit in the hole, it swells to become extremely tight in the hole. The metal being joined also deforms in response, especially if near an edge (bulging is sometimes seen). I haven't sectioned a rivet yet but even .010" growth on dia accounts for a fair proportion of the extra metal. One assumes small dimensional variance in the formed head itself would account for the rest.

Conclusion supports what I have usually discovered: that 1.5 d is never enough to form a properly proportioned head. The actual length protruding is the subject of experiment and is influenced by many factors from fit in the hole to surface finish to plate thickness to hole dia to rigidity of the plates to various other matters.

Much longer than 2 d becomes difficult to form on centreline so I would start there and tune length accordingly.

[92220]

I have just done a load of riveting on my tender frames. I machined down a piece of steel bar to the same total thickness as I wanted to rivet. I drilled holes the same diameter as I was going to use in the frames, and cut rivets to different lengths (within a tolerance of +/- 0.005") with my Arrand Rivet Cutter and formed rivets in the bar until I found the right length for the head size I wanted to produce. The tolerance on length is actually quite critical if you want all the rivet heads to look the same. It did work out at around 1.1/2 times the rivet diameter, but it is worth experimenting because a lot depends on the depth of the rivet snap, and ultimately, the height if the formed rivet head. Some rivets have shallower heads so need shorter rivets to form the head. Experiment is the only way you will get what you want. Specified protrusions are only a starting point.
The Arrand cutter is basically the same as Suctionhose describes, but it is 2 hardened, and ground, steel plates, with lots of different size holes in both plates, sliding across each other inside a frame and the cutting is done by using the vice to close the cutter and cut the rivet. To get the correct length, a collar is turned up with a hole for the rivet and parted off the length required minus 0.125", which is the thickness of the cutter blade. The collar is fitted over a longer rivet which is inserted through the holes in the plates. The vice then squeezes one plate across the other and shears the rivet at the sliding face. A simple thing to make up and use.

[uuu]

We've been riveting in the Pumphouse lately. Into countersink holes, so a little bit more forgiving - we can file off the excess, or add a dob of filler if it's a bit under. Complicated by the variable thickness of the casting being addressed, so calls of "next one a bit longer, please".

What we have developed is a preference for making the countersink with a Slocombe drill at 60 degrees, rather than the usual 90. The deformation required is less, and the hole seems to fill more easily. We're not lacking in compressive force, using a fly-press.

Wilf

[92220]

Wilf.

You may be interested to know that back in the days of steam, British Railways used 60 degrees inc. for countersink rivet holes - Ref SL/DE/22052 Frame Drilling Sketch for 9F locos, so you are in good company!! You are right as to the reason. It's much easier to fill a 60 degree c'sk than 90 degrees.

Bob

[johnthepump]

We should be riveting in the Pumphouse this evening, the angle iron frame stiffeners all drilled and waiting. Wilf should be here in the next half an hour and no doubt he will machine the rivets to length while I set them on the fly press.
John.

[johnthepump]

Dec 7, 2016 19:03:39 GMT johnthepump said:

We should be riveting in the Pumphouse this evening, the angle iron frame stiffeners all drilled and waiting. Wilf should be here in the next half an hour and no doubt he will machine the rivets to length while I set them on the fly press.
John.

As it turned out the rivets didn't need shortening. The total material thickness was 0.315" and the rivets were 0.125" diameter, so using the notation that to set the head requires 1.5 x diameter,
0.125 x 1.5 = 0.1875". A half inch long rivet gave 0.5 - 0.315 = 0.185".

There were 31 rivets to set in each frame and with three of us in the workshop last evening the job went smoothly. I have posted a photo in From the Pumphouse thread.
John.

[suctionhose]

John,
How's the formed head compare in size to the preformed head? Hard to see from your picture but are they smaller?

[uuu]

I'll nip in...

The first trial delivered a poor head shape - it wasn't as high, and too wide. So John modified the tool with a grinder to make the pocket a little deeper and narrower. Second trial was better, but left a cylindrical section the tool wasn't closing enough to fill the pocket - so a few thou was taken off the rim of the pocket in the lathe. Then the result was good - both heads about the same.

Wilf

[johnthepump]

Hi Ross,
To a great extent Wilf has answered your question, but there was a bit of history to the snaps we used. I bought a No.2 flypress in the mid 1970's and it came with a home made 1/8" rivet setting tooling. They had been used to make a small traction engine tank and bunker etc. Over the years they have got a little bruised and mishapened.
When we set the first rivet neither of the heads looked very nice, so I set about tidying them up, I found they weren't very hard and after trying little grinding stones in the Drexel type unit. In the end a 6 mm. Ball nosed carbide endmill did the job, both the top and bottom sets had the same treatment and the formed and fitted the heads to the same size. Hope this helps.
John.

[Shawki Shlemon]

I can only say from my experience in aircraft industry , that is 1.5 times the diameter , pull the plates together before riveting , I have a snap with a hole in it to drive the two plates hard together , then rivet .

[suctionhose]

Dec 9, 2016 8:24:19 GMT Shawki Shlemon said:

I can only say from my experience in aircraft industry , that is 1.5 times the diameter , pull the plates together before riveting , I have a snap with a hole in it to drive the two plates hard together , then rivet .

Hi Shawki,
Agreed for thin plates like aircraft sheet metal. As I said, the volume of a "standard head" works out to be equivalent to 1.5 x shank dia with out allowance for any to disappear in the long hole of thicker plates. I first noticed the difference repairing traction engine boilers where a rivet through the foundation ring was 5 or 6 inches long to start with. Quite a lot disappeared down the hole before the head began to form.

All that aside, I spent a day mucking about with snaps, much like Wilf and John have been doing this week, with the following results. The red circle is around a "manufactured head" and shows comparison with two of the heads I have formed immediately to the right. The four in a row is a completed spoke.

I have made the heads a bit smaller than those first reported and that in turn brought the shank protrusion closer to 1.5 x dia. I'm happy about that because I don't have cut a longer rivet back any more. A standard length does the job.

However, you can see the resulting head is slightly smaller than the manufactured head (red circle). Not entirely satisfied but I'll probably live with it!



.