Thread checking application

I have a machine that I have to design and build that will check the treads on a nut. The nut gets welded to a part and the whole part will then need to be loaded into this new machine. The nut is being checked for thread condition. I'm more concerned with deformation or weld spatter.

I'm currently looking on the web for a manufacturer of a couple of air operated thread chasers. What I'm looking for is something like a nut runner that will give an over torque feed back signal.
I would need two for my application because there is a RH and LH part that will be done at the same time. Something with good RPM is a must, do to the volume of parts being checked. I say air operated because I have to keep the cost down as much as possible.

Any ideas would be helpful.
Tim

Check ARO: http://www.arozone.com/Tool/ind_tools_screwdrv_nut.html

or Cleco: http://www.cooperpowertools.com/brands/cleco/index.cfm

or GSE: http://www.gsetechmotive.com/

I would look at a machine vision system to see if it could work for your application. I know many can do weld inspections and more. This would eliminate adding more moving components.

This link shows a Cognex system that has to inspect threaded holes:
http://www.cognex.com/pdf/industry/EyeOnQuality.pdf

Instead of measuring the torqe can you limit the torque, then use a prox to detect if the thread chaser went all the way through? That eliminates an analog input. If it stalls, then the chaser wont go all the way through and trip the prox. You wont know the actual torque required for each nut, but you will have a go/no-go test at some specified max torque.

What I hear you saying is...

You are welding a NUT onto a part and you want to verify that the threads IN THE NUT are good.

My suggestion is as follows... for whatever driver you might finally end up using...

The driver must be able to be properly positioned.

The part should be placed, or exist, in a jig for testing.

If the nut is welded onto a flat-piece, where you can NOT see through the nut and flat-piece to light, then the thread-tester-bit should be a "bottom-thread" tap.

If the nut is welded onto a flat-piece, where you CAN see through the nut and flat-piece to light, then the thread-tester-bit could be either a normal tap or a "bottom-thread" tap.

The thread-tester should be fixed in position or be able to be placed into the proper position for testing.

Once aligned for testing, the thread-tester should move to the part (or the part moved to the tester), pneumatically until the thread-tester and part are together and the psi driving them together is "X".

There must be a position sensor that can read the position of the thread-tester or the part... you need to know the relative positions.

After bringing the two parts together, and when the psi is "X", rotate the thread-tester SLOWLY in the direction opposite to the nominal nut-thread direction until there is a sudden decrease in psi or a sudden decrease in the relative position between the two. You need to have a good means to detect that sudden change in psi or position. The slower you run in the counter-direction the easier it will be to detect the change (either psi or position).

Then, at what ever minimum speed you absolutely must have, drive the bottom-thread tap forward (CW or CCW, as required) until psi reaches some max limit for a particular length of time.

In some cases, the tap can clean the thread but it needs a bit of time to do so.

When the max limit in psi is reached and held of a particular time, read the position sensor to determine whether or not the thread-tester is at the actual bottom.

If the position sensor indicates "at bottom" then the part is good. Otherwise, the part is NOT good.

You may also want to look at the actual process and prevent the problem in the first place. If you have time to chase the threads, then maybe you could screw the nut onto a bolt that is actually part of the fixture before welding. Then the fixture would be rotated outwards and the nut would be free of splatter.

You might also consider resistence welding instead of wire. Resistence welding requires no wire or rod, and there is no splatter or arcing. You can stand right there and watch it weld with no eye protection. Welding a nut to a metal part seems like the perfect application.

You might also consider resistence welding instead of wire. Resistence welding requires no wire or rod, and there is no splatter or arcing.

Click to expand...

I just wanted to add that this is not necessarily true. We have around 30 nut welders in house (all resistance welders) and this is a common problem. I don't have a lot to add to this because we just have manual checks. Every startup and at certain intervals during the shift the operators check the threads with a thread gauge. We have made pins for the operators which is slightly smaller (.005") than the outside diameter of the threads so the operator can quickly check the threads more regularly.
Of course this does not work for deformed threads just weld spatter on the outside of the threads.

Instead of measuring the torqe can you limit the torque, then use a prox to detect if the thread chaser went all the way through?

Click to expand...

We have a similar application for inserting bolts. We locate the part by the nut with a spring-loaded plunger and run down the bolt until the plunger activates a prox, which kills the air to the gun (We use this method because we have to leave the bolt a little loose for our customer). It seems like this would work for you just have the gun reverse instead of shut off.

Please keep us posted. I would be interested in hearing your solution.

I've suppled many of these to Ford, GM, and their tier 1,2 suppliers looking for thread deformation from mig welding. If you use a tap or die (something with a cutting edge) you may cross thread the existing thread and not detect it (huge no-no). You have to use a go-gage thread check ( min material ). Mount it in a nut-ruuner like a GSE, FEC-USA, etc with a holder that allows some float. Machines I've delivered using this approach have lasted a long time and few service issues compared to home brew methods. Your life will be a whole lot easier and you can get on with the next project.

PS: the nut runners will allow you to fault on number of turns (angle), max torque, etc.

Tlvaun, that's interesting. The parts we were welding were polished and could tolerate no splatter whatsoever, which is why we used resistence welding (we already used wire welders elsewhere). Unless I cranked the amperage way up and fired the cycle before full contact was made, I can't even imagine how I would get it to arc. It must have something to do with the part properties.

We usually weld 8 or 10mm nuts onto about 16 gauge steel. The nuts all have 4 projections on them that gets melted into the part. A lot of problems usually come when we have a very high torque requirement on the nut. It seems that there is a fine line between getting the correct current, and pressure settings to meet the requirements or heating it too fast and blowing some of the projection inside the nut.


BTW Tim you never did say if in your current process you mig or resistance weld the nuts.

Alaric said:

Instead of measuring the torqe can you limit the torque, then use a prox to detect if the thread chaser went all the way through? That eliminates an analog input. If it stalls, then the chaser wont go all the way through and trip the prox. You wont know the actual torque required for each nut, but you will have a go/no-go test at some specified max torque.

Click to expand...

I vote for this idea. Simple and streight forward. A small motor with a torqued controled drive driving the screw. If the screw hasnt reached a predtermened depth in a preset amount of time..REJECT.

The other consideration is the possability of bird$hit on the top of the nut preventing the screw from even starting. This will result in no high torque feedback (as the screw wont even get into the thread, but just ride on top of it), so with this setup the timer will be the deciding factor if the threads are damaged inside the nut or right at the top.

Your answer is Nycote coating on the threads. Nycote coating prevents the weld spatter to go on the threads. check www.Nylok.com