True Position of External Thread

[ddevaney1985]

I am currently having a problem figuring out how to measure the true position of a thread on a pin (5.6250-12 UNJ-3A) in relation to the main o/d. I obviously cant make a mating part to measure over as any tolerance in the thread and mating part would create slack.I need a repeatable way to measure this accurately. Our CMM software can chase a thread if we are given a 3D model with the thread created, however the models our company creates doesnt actually model the thread, a note is used and therefore time is saved in creating many models. Any ideas?

[MBG]

I am currently having a problem figuring out how to measure the true position of a thread on a pin (5.6250-12 UNJ-3A) in relation to the main o/d. I obviously cant make a mating part to measure over as any tolerance in the thread and mating part would create slack.I need a repeatable way to measure this accurately. Our CMM software can chase a thread if we are given a 3D model with the thread created, however the models our company creates doesnt actually model the thread, a note is used and therefore time is saved in creating many models. Any ideas?

Prints? What are the datum call outs from the tread? Is maximum material call out in there?

Also? On your Cmm are you using a round probe?

[ewlsey]

Been here with an internal thread. We programmed the CMM to take hits in a spiral pattern following the thread pitch. That worked OK, but was still taking bad hits.

We ended up buying a dedicated gauge from AG Davis. $5000 later, we found out they were good .

[ddevaney1985]

We use Nikon CMM's with Renishaw probing kits (SM25 etc etc). The callout is a true position of 0.005" Dia to Datum A (which is the main o/d on the same c/line with the thread).
Ewlsey - so when you cmm checked around the thread, was it not repeatable and accurate? Are there qaugings for external threads?

[ewlsey]

I was thread milling the thread (it's similar in shape to a large SAE port). The thread had a run out call to a feature deeper inside the cavity.

Since I was milling the thread, I was confident in the run out. However, our CMM was inconsistent. We went so far as to contact the customer to see what they recommended. They were using a laser scanning head on their CMM. That was not an option for us, so we bought the gauge. It uses a threaded body with a type of clamp to center and hold it on the thread. Then there is an indicator that sweeps the beveled lip of the SAE port. It works pretty well.

I think one could be made for an OD thread.

[Gordon B. Clarke]

I am currently having a problem figuring out how to measure the true position of a thread on a pin (5.6250-12 UNJ-3A) in relation to the main o/d. I obviously cant make a mating part to measure over as any tolerance in the thread and mating part would create slack.I need a repeatable way to measure this accurately. Our CMM software can chase a thread if we are given a 3D model with the thread created, however the models our company creates doesnt actually model the thread, a note is used and therefore time is saved in creating many models. Any ideas?

I'm at a loss trying to understand what you want to do and why.

The pitch diameter tolerance (and no matter what you make there will always be a tolerance) is the main factor with slack. Pitch is also a factor but length of engagement plays into that.

As tolerance class 3 is the pitch diameter tolerance with the smallest tolerance then slack will be minimal. No slack in a thread can be achieved but that's another story

If it is really critical that you know then there are companies/laboritories that could measure what you want measured. It'll certainly be a whole lot cheaper than buying a machine that can do it.

I've seen a profile measuring machine made by Hommelwerke that can measure just about everying with extreme precision (0.001mm) but it costs both an arm and a leg.

Gordon

[Eric M]

ddvaney - Have you tried using a cylindrical probe on your CMM? You're less likely to make bad hits, since the cylinder will only pick up the crest of the thread.

[The real Leigh]

I'm at a loss trying to understand what you want to do and why.

Hi Gordon,

This is how I interpret the problem:

Determine whether or not the centerline of the thread is concentric with the centerline of the part, and if not, how far out is it?

Think of a shoulder cap screw... Determining whether or not the screw thread and the shoulder are concentric.

Or perhaps that's completely wrong. We need input from the OP.

- Leigh

[ddevaney1985]

Basically, im trying to measure the true position of the thread, off the effective diameter (Per ASME Y14.5 2009 Geometric Dimensioning & Tolerancing) in relation to the O/D centre line and it must be within a diameter tolerance of 0.005". Aerospace first articles require that i output the position as drawing states, even though i know that its produced at the same time it will be accurate - however i need to output a result and a repeatable one due to its KEY CHARACTORISTIC status. Sadly for some reason i cant upload a drawing of it as example.

[The real Leigh]

Is my comparison with a shoulder cap screw in post #8 correct?

- Leigh

[ddevaney1985]

Yes, thats a good example but on a larger scale, same principle - the thread needs to be on the centre of that shoulder within the 0.005" diametric tolerance.

Darren

[Gordon B. Clarke]

Hi Gordon,

This is how I interpret the problem:

Determine whether or not the centerline of the thread is concentric with the centerline of the part, and if not, how far out is it?

Think of a shoulder cap screw... Determining whether or not the screw thread and the shoulder are concentric.

Or perhaps that's completely wrong. We need input from the OP.

- Leigh

That is how I see it too but it seems more like a theoretical task than a practical one. If the thread OD is turned on a lathe and the thread machined (without moving anything) I don't see how there can be a difference in a common shared centerline.

If a thread is machined using a tap or die then there could be a discrepancy but with a tolerance class 3 a die or tap is inadvisable if a common centerline is crucial. A thread with the diameter given in the OP will certainly be machined by a threading tool.

I've noticed some call the pitch diameter "effective diameter" but they are both the same thing. Pitch diameter is the more correct of the two.

Bolt Science - Basic Terminology related to Screw Threads

My opinion is that some newbie in the drawing office has been bitten by the geometric dimensioning and tolerancing bug and is applying it to as much as possible.
GMT is excellent - when used for the intended purpose.

I realise that my post isn't helping the OP any but if proof is required then a surface projector is necessary or the much more expensive solution - a profile measuring apparatus. Those I believe start at around $100,000.

What would I do? If the customer must be satisfied then find a company that has a surface projector and see if you can use it for about ½ an hour or pay to have them do it for you.

0.005" gives quite a bit of leeway To be that far off would require extra effort.

I'm really fortunate in that I have a large international (but Danish owned) company 5 minutes drive from me with just about every imaginable measuring machine. I do them favours and they do favours for me. You might want to try the same approach with a company near you.

Gordon

[CarbideBob]

My first go to here would be an optical comparator if the part will fit.

A machine vision head on a cmm would also work.

Probing this on a cmm is going to be very iffy and will need a lot of data points for a result that does not bounce around but if you are good enough at programming one it can be done.
It will not be fast.
If you don't have a lot of custom cmm programming experience contact the maker of your machine and see if their application department will whip up a program for you at a reasonable cost.

On high volumes my customers would purchase a dedicated video inspection system with 4 cameras or "image split" optics. Not cheap.

The use of these is common for holes but I have not see ones for OD threads. A thread grinder could make you a special but it would need to be "split" to lock onto your threads.
The Gage Store - Measurement Tools and Accessories

[The real Leigh]

My opinion is that some newbie in the drawing office has been bitten by the geometric dimensioning and tolerancing bug and is applying it to as much as possible.

If I had a dollar/Euro for every time that's happened, I wouldn't be sitting here typing on an internet forum.

- Leigh

[The real Leigh]

Question to the OP:

Is this thread really 5.6250-12 UNJ-3A, or is it supposed to be 0.5625-12?

If there's a typo I'll change it.

- Leigh

[Gordon B. Clarke]

Question to the OP:

Is this thread really 5.6250-12 UNJ-3A, or is it supposed to be 0.5625-12?

If there's a typo I'll change it.

- Leigh

Hmmmm I believe it's a typo. 0.6250-12 is a standard UNC thread and the only difference between UNJ and UNC is a small difference in the radius.

On the other hand 5.625-12 is a UN thread.

Will the real diameter please step forward?

Gordon

I don't know if this is helpful to anyone?

http://www.ring-plug-thread-gages.com/ti-UNJ-vs-UN.htm

http://www.f-m-s.dk/1.02.pdf

[The real Leigh]

0.6250-12 is a standard UNC thread

So is 0.5625-12 (9/16" for the fractophobes).

But yes, 0.625 would be another option.

- Leigh

[Bruce Nelson]

Gordon:
It would seem to me that with your expertise at making gaging components with a wire edm, you could make a set of four gages with 12 tpi that could be assembled in engagement with the threads around the part to be checked for true position. If these gage components were made out of 1/32 inch thick flat tool steel and a cage made with 4 slots to hold these gage components in position when checking the position of the thread pitch diameter in relationship with the theoretical true position.

If the gage components were made a specific width, they could be assembled around the thread to be measured for position of the pitch diameter, using the slotted cage, and an O-ring to keep the gage components in contact with the flanks of the thread. If this were used in conjunction with a CMM probe you could determine whether or not the thread pitch diameter was within the tolerance of the true position.

On further reflection, a cage with 3 slots in it may be more appropriate, because most CMMs will output a diameter with 3 hits of the probe.

Lord Byron

[Gordon B. Clarke]

Gordon:
It would seem to me that with your expertise at making gaging components with a wire edm, you could make a set of four gages with 12 tpi that could be assembled in engagement with the threads around the part to be checked for true position. If these gage components were made out of 1/32 inch thick flat tool steel and a cage made with 4 slots to hold these gage components in position when checking the position of the thread pitch diameter in relationship with the theoretical true position.

If the gage components were made a specific width, they could be assembled around the thread to be measured for position of the pitch diameter, using the slotted cage, and an O-ring to keep the gage components in contact with the flanks of the thread. If this were used in conjunction with a CMM probe you could determine whether or not the thread pitch diameter was within the tolerance of the true position.

On further reflection, a cage with 3 slots in it may be more appropriate, because most CMMs will output a diameter with 3 hits of the probe.

Lord Byron

Now there's a coincidence I'm just about to draw up an idea I have as to how pitch can be measured. What I'm interested in first is how many "inserts" would be necessary to cover the range 05-8mm/48-3 TPI. What I have in mind (it always looks so simple when still in my head LOL) would be able to measure over one or umpteem pitches and both externally and internally.

At the moment I'm assuming the greatest accuracy will be measuring over several pitches and dividing by the number of pitches and comparing that result to just measuring over a single pitch.

Just to clear things up - I am referring to length measurements and not diameters

Apart from a thread OD (d) and an internal bore diameter (D1) which are both easy to measure, I have inserts for d2, D2, d1 and D. Pitch would seem the next logical progression. When I have more I'll post it in here as I can't imagine pitch measurement ever being a "seller".

As to my "letters and numbers" for thread diameters then for those that don't know this is what they stand for:

External thread
Major diameter = d
Pitch diameter = d2
Minor diameter = d1

Internal thread
Major diameter = D
Pitch diameter = D2
Minor diameter = D1

Gordon

Bruce, I've just realized that my thoughts were wrapped around the other (now locked) thread re pitch and has nothing to do with this particular thread. What you suggest can almost certainly be made but would probably be considered too expensive by the OP.

[ddevaney1985]

The thread is infact 5.6250 - 12UNJ - 3A. I totally agree about the drawing office, they dont realise that sometimes its just not practical to measure these things, and i will know its good because as Gordon states, if they are turned at the same time then they will be concentric to one another! Thanks for all your input guys

Darren