Semi-tubular Rivets Made In-House are Splitting

Nerdlinger · Apr 17, 2020

Hi Everyone!

We have an application where we peen over a .156" diameter X .750" long solid rivet with a hammer. The rivet is steel, hardened to 40-42 HRC. I drilled a small hole in the end to make it semi-tubular and it formed over beautifully with a standard set. I got tired of modifying them by putting them in a speed lathe one at a time to drill out the hole so I made a bunch out of 4140 in a CNC and heat treated them but those split when set (that is, the portion that got formed/rolled over split.) I tried waxing them but no improvement. I imagine there is some difference between the material that the lifetime supply of solid rivets we have is made of and 4140 or it has something to do with the cold forming the solid rivets underwent. Does anyone know what a good steel alloy would be to make a semi-tubular rivet out of in a CNC lathe that can also be hardened to ~40 HRC?

Thank you!

Disclaimer - I am NOT a rivet guy so I apologize if my terminology is off.

Limy Sami · Apr 17, 2020

I am NOT a heat treat guy but;-

I'd be looking at the heat treat on the faulty rivets, ………..maybe normalise and re - heat treat by a different method.

TDegenhart · Apr 17, 2020

Several aspects. What was the steel that you used to make the hand made rivets, both composition and physical properties. Why did you go to a 4140 and heat treat? What hardness did you heat treat to? How did you curl the first parts, what type and shape tool? Did you impact(hammer) curl the second set as you did the first set?

Where I worked, there were millions of semi-tubular rivets used in assemblies. An older and experienced tool maker told me that the shape of the curling tool will make all the difference between a good and a bad curl. At that time I wasn't particularly interested in the details, but that tidbit has always stayed with me.

Tom

Nerdlinger · Apr 17, 2020

TDegenhart said:
Several aspects. What was the steel that you used to make the hand made rivets, both composition and physical properties. Why did you go to a 4140 and heat treat? What hardness did you heat treat to? How did you curl the first parts, what type and shape tool? Did you impact(hammer) curl the second set as you did the first set?

Where I worked, there were millions of semi-tubular rivets used in assemblies. An older and experienced tool maker told me that the shape of the curling tool will make all the difference between a good and a bad curl. At that time I wasn't particularly interested in the details, but that tidbit has always stayed with me.

Tom

Hi Tom! The original batch of solid rivets that I modified and worked great are so old we have no paperwork on them but the prints specs "1038, 1041, 4037 or approved substitute." The guy that wrote that spec is long gone so I just kind of randomly picked 4140 just because I knew it could reach the 42HRC spec. I used the same tool setter on both 1) the solid rivets that I modified which worked and 2) the ones I made from bar stock that did not work. Here is a picture of the setter...it's a Chicago Rivet T978 anvil. Maybe the rivet is too small for it? It's .154OD X .109ID??

DanielG · Apr 17, 2020

The first thing I'd check is hardness. What is the measured hardness of the old rivet, both on the surface and at the centerline. How does this compare to the new one? It's possible the old one had the required 42 hardness on the outside, but a softer core.

sfriedberg · Apr 17, 2020

Yeah, you need a lot of ductility to form rivets properly. It's quite possible that your new rivets are too hard, or have been through-hardened when the old ones were only case-hardened.

thermite · Apr 17, 2020

sfriedberg said:
Yeah, you need a lot of ductility to form rivets properly. It's quite possible that your new rivets are too hard, or have been through-hardened when the old ones were only case-hardened.

"In the curl" they wudda been WORK hardened like a bastid, as well.

You don't really want a Cr-Ni-Mb steel in there at all, actually.
Vanadium, or even a touch of Manganese, if anything.

"1038, 1041..." were permitted? JF try those or their "cousins".

4140 is "more different than" 4037 than they are, one from another.

Need "approval"? Get your samples up to snuff, submit the paperwork for them, problem solved.

Expect to get approval, too, if it's reliably repeatable RESULTS, "modernized", the buyer needs.

Phil in Montana · Apr 17, 2020

I cant think of a worse choose in steel for rivets than 4140ht....Phil

thermite · Apr 17, 2020

Phil in Montana said:
I cant think of a worse choose in steel for rivets than 4140ht....Phil

LOL!

Shall we "try harder?"

How about M42?

CalG · Apr 17, 2020

What can be said?

4000 series steel is WRONG.

Do some homework. Machinery's Handbook would give you several better choice.

So would a steel supply house.

kustomizer · Apr 17, 2020

Try putting a thin washer on it before you peen it, perhaps it is just a tad too long

cyanidekid · Apr 18, 2020

Super Common mistake is to think a number defines the steel. No it doesn’t.

it only defines a range of analysis, and a wide one at that. A range of alloying constituents, ABSOLUTELY NOTHING about the work processing of the metal, it’s crystal structure or freedoms from critical flaws.

Is it “cold heading” grade? But that’s all beside the point.. Hardened 4130 and then you wonder why it’s splitting? Basic materials science says if you need the hardness and the ability to cold work, you start with a more malleable crystal structure.
select a cold heading grade, derive needed hardness from cold work, and have enough remaining deformability to achieve the set. At least I think so, something like that..I think you could still do this in “4130” if it was cold heading grade 1/8 hard or so cold worked. Proof of this is the billions of SHCS made annually.

Nerdlinger · Apr 18, 2020

Okay I’m not hearing a lot of support for 4140, or at least the grain size and other non-alloying-element-related attributes of the specific batch of steel I used. Does anyone have any recommendations as to a more-appropriate material spec? I’m kind of surprised the solid rivets I drilled out formed so well being 42HRC.

Do you think (or know if?) it is even possible to machine rivet wire and have it perform like a cold-headed rivet, as far as “formability” goes? I know cold heading is faster but does it also yield necessary mechanical properties one cannot achieve with cutting tools, like the ability to be hardened yet still be rolled over as a semi-tubular rivet does?

On Monday I will check the hardness on the old solid rivets.

Thank you!

cyanidekid · Apr 18, 2020

Nerdlinger said:
Okay I’m not hearing a lot of support for 4140, or at least the grain size and other non-alloying-element-related attributes of the specific batch of steel I used. Does anyone have any recommendations as to a more-appropriate material spec? I’m kind of surprised the solid rivets I drilled out formed so well being 42HRC.

Do you think (or know if?) it is even possible to machine rivet wire and have it perform like a cold-headed rivet, as far as “formability” goes? I know cold heading is faster but does it also yield necessary mechanical properties one cannot achieve with cutting tools, like the ability to be hardened yet still be rolled over as a semi-tubular rivet does?

On Monday I will check the hardness on the old solid rivets.

Thank you!

go open the drawer and pull out a 1/4-20 socket head cap screw. that is cold headed, and the threads are rolled on at room temperature also ("cold" for our purposes here).

its probably harder than 42. again, this is basic stuff, it's not the alloy, its not the hardness, its not even the grain size (though very important), its the crystal structure.

you have seen a "phase diagram" for iron-carbon somewhere in your travels I'm sure. the crystal structure of the high temp phase, which you "freeze" in place by quenching, is resistant to being plasticly deformed. that is one of the most useful things about steel, but NOT what you want in a rivet.

hardness is a RESULT not a process. It doesn't only come from heat treating, as anyone who has spun a dull drill in stainless can attest.

machinists tend to make the mistake of thinking the best (only?) way to make a strong part is cut and heat treat. actually, a cut screw is FAR, FAR inferior to a rolled thread and a cold headed part. this is nothing new, really basic stuff. absolutely no way modern machines from 200 year old steam boilers to spacecraft could be possible without cold headed fasteners.

just a little tip, they don't usually start with quenched and tempered steel

TDegenhart · Apr 18, 2020

Check out the ductility ratings of the various steels.

Tom

cyanidekid · Apr 18, 2020

TDegenhart said:
Check out the ductility ratings of the various steels.

Tom

the ductility rating you will find "of the various steels" is a rough approximation. the actual workable ductility of a given piece of steel isn't determined by some 4 digit number in a table, but a host of factors pertaining to that actual piece, such as the sulfur and phosphorus content, the other alloy constituents, which can vary quite a bit and still be "within spec", the processing of the shape both thermally and mechanically, the crystal structure (yup that again!), the condition of the surface, the presence of flaws, etc etc...

no high performance fastener manufacturer would even consider just grabbing a piece of "4130" off the shelf.

who is calling for "42RHC" anyway, and how is that inspected, and on what area of this feature? to what depth? why are you trying to make this in house? can't you source this? just wondering.

Nerdlinger · Apr 18, 2020

So are you saying the resulting crystal structure from cold forming is likely what gives the solid rivets I have the combination of 42HRC hardness(strength) and ductility that allows them to be drilled into fully-formable semi-tubular rivets?

Also, since I’m a little short on cold heading machines, is it possible to achieve that same crystal structure via heat treating (the only hardening process I really have access to that I know of)?

The spec is simply called out on the part print by an engineer that is no longer alive. We got them in many years ago so inspection records are gone. We would likely check them on the HRC scale directly or 15N scale and convert to HRC, coming in from the side of the shank radially. I am trying to make these in house because volumes are too low to buy from a rivet mfg (~5,000/year). This is for our own products we manufacture, by the way. Thanks!

NewGunPlumber · Apr 19, 2020

Not sure on the material but a friend had a boiler repair done for an old traction engine and apparently boiler rivets need to be formed and not machined to prevent cracking?

Street · Apr 19, 2020

different rivets done completely differently on a boiler....

the machines we use sort of roll the tool around to form the rivet, it only expands the head around .5mm -0.8mm or so depending on diameter to be ok.

i don't have a material spec or heat treat to help you though.

cyanidekid · Apr 19, 2020

Nerdlinger said:
So are you saying the resulting crstal structure from cold forming is likely what gives the solid rivets I have the combination of 42HRC hardness(strength) and ductility that allows them to be drilled into fully-formable semi-tubular rivets?

Also, since IÂ’m a little short on cold heading machines, is it possible to achieve that same crystal structure via heat treating (the only hardening process I really have access to that I know of)?

The spec is simply called out on the part print by an engineer that is no longer alive. We got them in many years ago so inspection records are gone. We would likely check them on the HRC scale directly or 15N scale and convert to HRC, coming in from the side of the shank radially. I am trying to make these in house because volumes are too low to buy from a rivet mfg (~5,000/year). This is for our own products we manufacture, by the way. Thanks!

OMG!! You are finally starting to get it!!

The answer to the first paragraph is You are half right. It’s not that you change the crystal structure by cold working, it’s that you DON’T. You make it harder from cold work WITHOUT changing to a brittle phase.

You answered your Own question from paragraph 2 in paragraph 1 , in other words NO.

FORGET HEAT TREATING THE PARTS! :wall:

Are you sure this isn’t available as an off the shelf item? How many rivet manufacturers have you contacted? Tubular end rivets are super common.

If really not a stock part, why aren’t you just buying rivets and drilling them? Wait, Isn’t that what you were doing before? Lol!

Semi-tubular Rivets Made In-House are Splitting

Stainless

Diamond

Diamond

Stainless

Stainless

Diamond

Diamond

Stainless

Diamond

Diamond

Diamond

Titanium

Stainless

Titanium

Diamond

Titanium

Stainless

Aluminum

Hot Rolled

Titanium

Similar threads