Welding two nuts inline- how to w/o binding?

[tarbellb]

Is there a method to achieving this fabrication design without them binding up?

I've been seeing others here like @Beerhippie and more hitting this same issue

My latest example was my welding table build, for the stabilizers I used this design. And while I tried everything I got mixed results. Two turned out pretty good, one needed cycling, and one needed full blown rethread and still isn't great?


So is there a surefire way to incorporate this or is it a flawed design?

table has 3/4x10tpi all thread not acme thread

 

[The Cobbler]

they get distorted from welding, a bottoming tap would be the easiest to restore I think

 

[Beerhippie]

In my experience, no. The alignment needs to be too perfect for the world I live in.

A good way to get more thread engagement is to use coupling nuts instead of regular nuts.

You could also drill out the upper nuts to just exactly the OD of the all-tread to add stability without the threads binding.

 

[Beerhippie]

they get distorted from welding, a bottoming tap would be the easiest to restore I think

If you could find one long enough to go through both bolts at the same time....

 

[The Cobbler]

find one long enough to go through both bolts

I missed the second nut on the top.

 

[Beerhippie]

I missed the second nut on the top.

Aye, there's the rub.

If those two nuts aren't clocked to withing one micro-RCH of angle of each other, they bind.

 

[LopezBart]

If you use a hot dipped galvanized nut on the top (and grind off the galvanizing before welding), it will be easier - those have a much looser thread tolerance. That's probably the quickest fix; obviously a design where the nut retains some freedom to rotate while still constraining side to side motion would be less finicky.

 

[whateg01]

Why do you need nuts at top and bottom? The threads of one should have plenty of strength to support that.

If I had to have that, I would back the top one off just a smidge so as the weld cools and pulls them towards each other, there's a bit of slop for that to happen.

But I wouldn't do that at all. I'd just use a single nut at the bottom.

 

[Hohn]

Is there a method to achieving this fabrication design without them binding up?

I've been seeing others here like @Beerhippie and more hitting this same issue

My latest example was my welding table build, for the stabilizers I used this design. And while I tried everything I got mixed results. Two turned out pretty good, one needed cycling, and one needed full blown rethread and still isn't great?


So is there a surefire way to incorporate this or is it a flawed design?

table has 3/4x10tpi all thread not acme thread

I'd suggest a design change. Two nuts both tied to the same axis on a common threaded rod is a design flaw, IMPO.

With two nuts both tied to the structure, only one of them can actually carry load anyway. One of the nuts is carrying axial load and the other (even if perfectly executed) is at best just locating your all-thread within the rectangular tube section.

You're trying to use one nut as a bushing. You need a bushing or a bearing. But bushings need smooth shafting, not threaded rod.


I'd think it would be preferable to transition from the all-thread to smooth shafting.
With smooth shafting, you can use a plate mount bearing on an end cap of your rectangular tubing.

McMaster-Carr

McMaster-Carr is the complete source for your plant with over 595,000 products. 98% of products ordered ship from stock and deliver same or next day.

www.mcmaster.com

With the portion of shaft that protrudes up, you can mount almost anything-- handwheel, pulley, or even a hex.

 

[whateg01]

I'd suggest a design change. Two nuts both tied to the same axis on a common threaded rod is a design flaw, IMPO.

With two nuts both tied to the structure, only one of them can actually carry load anyway. One of the nuts is carrying axial load and the other (even if perfectly executed) is at best just locating your all-thread within the rectangular tube section.

You're trying to use one nut as a bushing. You need a bushing or a bearing. But bushings need smooth shafting, not threaded rod.


I'd think it would be preferable to transition from the all-thread to smooth shafting.
With smooth shafting, you can use a plate mount bearing on an end cap of your rectangular tubing.

McMaster-Carr

McMaster-Carr is the complete source for your plant with over 595,000 products. 98% of products ordered ship from stock and deliver same or next day.

www.mcmaster.com

With the portion of shaft that protrudes up, you can mount almost anything-- handwheel, pulley, or even a hex.

Unless that table is getting raised and lowered 5 times every day, a simple hole through a piece of 1/4" is plenty. There's not going to be that much radial load on it if the nut is at the bottom.

 

[Hohn]

Sorry, a bit more spit balling....

Since the top nut isn't needing to bear any axial load, why does it need to be welded? Why nut use a four-sided nut, drop it INSIDE the rectangular tube end and use a cold bond (epoxy) or such so secure it?

You're really just using the top nut to keep the all thread from located within the tube. If you ground down a nut (or found one) that *just* fit inside the tube ID, it would not rotate and the only issue would be the clearance of the nut within the tube making a rattle under vibration. Does that matter?

 

[whateg01]

Sorry, a bit more spit balling....

Since the top nut isn't needing to bear any axial load, why does it need to be welded? Why nut use a four-sided nut, drop it INSIDE the rectangular tube end and use a cold bond (epoxy) or such so secure it?

For location it wouldn't even need that. It could float

 

[tarbellb]

Agreed on all these points fellas

Interesting point from @LopezBart about the lower tolerance galvy nut

I did realize that only 1 nut was needed for the load, but I don't have the tooling to hone out the diameter on the opposing nut...

Hind sight I wouldve run just one and let the all thread float above

 

[rancherbill]

Weld the bottom collar nut and install on the leg. Weld another collar nut for the top and let it cool. Put it on the threaded rod and do 2 tacks, go away and let it cool and come back and do 2 more tacks. Let it cool and if you want more you should be able to go wild.

Tension is important to get it to move later. Don't have it tight or loose the other way have the tension neutral while finish welding.

 

[tarbellb]

All 4 do operate as intended
Some just need more ugga chugga from the M18

 

[danielbuck]

when I weld up control arm tube inserts for vehicles, it's often a similar sized thread. I just insert a bolt in the threads with high temp anti-seize on the threads. let it cool and then remove the bolt. I've never had one bind up. I know that's not exactly the same.

 

[Beerhippie]

Agreed on all these points fellas

Interesting point from @LopezBart about the lower tolerance galvy nut

I did realize that only 1 nut was needed for the load, but I don't have the tooling to hone out the diameter on the opposing nut...

Hind sight I wouldve run just one and let the all thread float above

You don't have a drill and bits? It ain't high-tolerance work--just making a nut into a loose bushing.

Weld the bottom collar nut and install on the leg. Weld another collar nut for the top and let it cool. Put it on the threaded rod and do 2 tacks, go away and let it cool and come back and do 2 more tacks. Let it cool and if you want more you should be able to go wild.

Tension is important to get it to move later. Don't have it tight or loose the other way have the tension neutral while finish welding.

If this can be done--and I'm sure it can--I have never gotten it to work.

A locking nut above the top nut--that you've now converted to a bushing--would do a better job.

For levelers that need to be adjusted frequently and quickly:



That's a coupling nut welded to the portable vise stand. The big nut on top is a knob. Q&D to adjust.

 

[T444e]

If you use a hot dipped galvanized nut on the top (and grind off the galvanizing before welding), it will be easier - those have a much looser thread tolerance. That's probably the quickest fix; obviously a design where the nut retains some freedom to rotate while still constraining side to side motion would be less finicky.

HDG nuts are tapped oversize to account for the galvanizing.

 

[Beerhippie]

If you use a hot dipped galvanized nut on the top (and grind off the galvanizing before welding), it will be easier - those have a much looser thread tolerance. That's probably the quickest fix; obviously a design where the nut retains some freedom to rotate while still constraining side to side motion would be less finicky.

Dissolve the galvanizing off with Muriatic or phosphoric acid. It's what I do when I need to weld on galvy. Takes a minute.

 

[Beerhippie]

Since some folks think this is something that can be done relatively easily, I'm throwing down the gauntlet: Weld two nuts on the ends of a 3" piece of pipe so a bolt can be threaded through both using fingers only.

If you pull it off, I'll shout you a beer at TG.

 

[T444e]

Sorry too far away, and I only stayed at a Holiday Inn last night. As I like to say, I only stick metal together at home.

 

[Hohn]

Since some folks think this is something that can be done relatively easily, I'm throwing down the gauntlet: Weld two nuts on the ends of a 3" piece of pipe so a bolt can be threaded through both using fingers only.

If you pull it off, I'll shout you a beer at TG.

Theres a reason that any manufacturing operation would cut the threads AFTER it’s welded. Welding on pre-machined parts and holding any kind of true position tolerance is all but impossible.

 

[imagineer]

When you reference "binding", do you mean the threads were buggered due to welding, or that the two nuts were not clocked correctly to allow the bolt to turn though both nuts at the same time?

A few years ago, I designed a purpose-built vise for a weld fixture, and to ensure the nuts were clocked properly, after I welded one of the nuts, I installed threaded rod with 2 nuts installed. Instead of tightening the 2nd nut to be welded to the assembly, I ran it the opposite direction, using the 3rd nut as a jam nut to lock the nut to be welded. Whereas this made the set up somewhat loose (maybe by 1/32") and I needed to be careful to keep it all aligned, when the welds on the 2nd nut cooled and shrunk, it pulled the nut toward the assembly just a little and ensured the threads were clocked. It worked well.

Another trick I use when welding a nut where the threads have to remain aligned, where possible, I'll use a machine tap in place of a bolt for holding the nut. The male threads on a machine tap are much more accurate than a bolt.

 

[rsanter]

It’s a flawed design IMO
if it was me I would use a coupling nut on the bottom for more thread engagement and then drill the top one to just be a close fitting sleeve

 

[KwikFab]

when I weld up control arm tube inserts for vehicles, it's often a similar sized thread. I just insert a bolt in the threads with high temp anti-seize on the threads. let it cool and then remove the bolt. I've never had one bind up. I know that's not exactly the same.

Did something similar at my first fab job when we made park benches and the sort.

Threading anything through never seemed to "bind" as others explained here.

 

[sz0k30]

!st choice is a coupling nut as mentioned in an earlier post. But if you think you really need to have all that much thread engagement, make your own. Buy some hex bar stock, cut to the length you need, drill & tap. I've done that (well I should say a buddy did it for me). Luckily I have a buddy with some nice shop equipment including a Bridgeport & a lathe!

 

[danielbuck]

I wonder if the issue is not the threads distorting, but the nuts actually being pulled towards each other slightly while cooling. It might be best to weld one of them up, let it fully cool, then tack the other nut in place with threaded rod aligning them, let that cool and then fully weld the other nut. I imagine that would cut this effect roughly in 1/2, maybe a little more.

When you thread a bolt in, does each nut spin freely, and only bind once you get the all-thread into both of the nuts?

 

[PugetDude]

Soak a galvanized nut in Citric Acid overnight, the threads will be loose enough that you can probably get it to work on the top side.
If not, run a tap through it a half dozen times using a drill, forwards and backwards with a bit of side pressure to open up the threads. As other have said you really only need a bushing; this will give you a very sloppy threaded one. Bottom nut is doing all the work anyway.

A thin jam nut would probably also work, fewer threads to bind up. Since you have already welded the legs up, just take half the width off the top nuts with a cutoff wheel in an angle grinder, then clean up the threads with a tap.

 

[WillyBoy]

Here are some pictures of the coupling nut style stabilizing feet that I use. These, shown, are on a frame I fabricated for a Powermatic 90. I also have the same style on the frame that holds two cabinet saws, the cabinet for a Logan metal lathe, and the frame I fabricated for a Clausing mill.






Unfortunately the black paint on the second photo doesn't show a lot of detail. There is a carriage bolt threaded through the nut upside down so the head of the bolt is on the floor when extended. After the bolt is in place, a nut is welded to the top end of the bolt. A long extension and an battery impact wrench work fine for running the bolts up and down when the machine needs to be moved.

I discovered right away that welding two faces of the coupling nut distorted the nut just enough to prevent a bolt from threading easily though it. Finding a socket that would slip over the end of a tap, and using an impact driver or drill to run the tap up and down through the coupling nut with some lube on the tap solved that problem.

It helps to sing the "grinder and paint" song while looking at the close-ups of the welding.

 

[Brianf60]

I’m not familiar with that tune. Could you sing it for me?

 

[Beerhippie]

Here's another problem: COE. As the tube with the nuts welded to each end changes temperature, it changes length--the longer the tube, the more it changes. With changing length, you effectively get changing clocking between the two threaded nuts.

 

[whateg01]

Here's another problem: COE. As the tube with the nuts welded to each end changes temperature, it changes length--the longer the tube, the more it changes. With changing length, you effectively get changing clocking between the two threaded nuts.

I happen to really like the old cab-over-engines. I would like to have an early 40s Ford COE!

 

[danielbuck]

Here's another problem: COE. As the tube with the nuts welded to each end changes temperature, it changes length--the longer the tube, the more it changes. With changing length, you effectively get changing clocking between the two threaded nuts.

I think that's probably the real issue here. I've welded a single nut on the bottom of alot of table legs, and in other places on alot of projects and almost never do I have one bind. I think the issue is the distance changing.

 

[WillyBoy]

Could you sing it for me?

Grinder and paint make a welder I ain't. LA-LA-LA-LA-LA

 

[T444e]

Here's another problem: COE. As the tube with the nuts welded to each end changes temperature, it changes length--the longer the tube, the more it changes. With changing length, you effectively get changing clocking between the two threaded nuts.

In my experience, hardware, like most steel products, is manufactured to the min. Meaning there is plenty of slop in the hardware. I'm not sure CTE is enough to affect thread clearance at the length the OP has.

I may need to try this for the hell of it. I'm sure I have some 3/4-10 ATR laying around.

OP, what is the face to face dimensions on the weldment? Did you use finish hex or heavy hex nuts? I'm know I have both of those laying around.

 

[PugetDude]

I have welded two nuts together before with decent results; threaded then onto a piece of all thread with a very small gap in between. Tack them together with a very small tack weld. Then use a pair of wrenches to turn them to see how they bind up; the small tack can be deformed one way or the other to get the threads clocked. Repeat on the other side, test again. Then weld the perimeter. I also run a tap through with my cordless drill when I am done.
(But, I also have an assortment of coupling nuts, a lot faster when I want a longer thread engagement on a weldment. Much easier )

 

[Muckin_Slusher]

Maybe after fabricating you could add something gritty (lapping compound?) to a sacrificial all-thread, then run it up and down a bunch while loaded to wear out the nuts a little. The part of the nuts causing binding should wear out fastest...

 

[tarbellb]

In my experience, hardware, like most steel products, is manufactured to the min. Meaning there is plenty of slop in the hardware. I'm not sure CTE is enough to affect thread clearance at the length the OP has.

I may need to try this for the hell of it. I'm sure I have some 3/4-10 ATR laying around.

OP, what is the face to face dimensions on the weldment? Did you use finish hex or heavy hex nuts? I'm know I have both of those laying around.

The face to face dim is 11.5"

Give it a shot, I'm interested to hear the outcome. What's the difference between a finish and heavy hex nut? Got mine from the local industrial supply, didn't spec anything besides grade 5.

 

[T444e]

Heavy hex nuts are an 1/8" wider across the flats than a finish hex nut (standard) I'll try next time I weld something up, hopefully this weekend. It's been awhile.

 

[kctgb]

Why do you need nuts at top and bottom? The threads of one should have plenty of strength to support that.

If I had to have that, I would back the top one off just a smidge so as the weld cools and pulls them towards each other, there's a bit of slop for that to happen.

But I wouldn't do that at all. I'd just use a single nut at the bottom.

That’s what I was going to say, you beat me to it.