Lining up pattern across threaded parts

[mitmeche]

Hi,

I need to align a grid pattern across two threaded parts. One part screws into the other and the resulting pattern must line up. This issue comes up rather frequently and I don't know the proper way to specify the thread on each part so my machine shop will do the right thing and the parts will come out like I want.

This is probably machine drawing 101 but although I have been drawing for a number of years I have never found a consistent answer/result in this situation.

Thanks for any help.

Cameron
Watertown, MA

 

[SlicerMan]

Design it to accomodate shims or washers of various thicknesses that can be chosen to make the pattern line up.
Or, pin it.

SM

 

[Carl Darnell]

I was taught to draw lines at an angle in the direction of left or right hand threads as the case may be but that is old school. I always note on the drawing with an arrow refferencing the thread what the thread is. For instance 1/2-13 RH or LH. Most the time if all the threads are right hand I don't note if it's left or right hand.

Is that what your asking??? Having read your post several times that is the only thing I could read in it.

 

[SlicerMan]

One part screws into the other and the resulting pattern must line up.
...I don't know the proper way to specify the thread on each part so my machine shop will do the right thing and the parts will come out like I want.

I just noticed you are an engineer.
No offense intended, but I bet the machine shop guys love you.

...I have been drawing for a number of years I have never found a consistent answer/result in this situation.

Because it's a pain in the axx and requires extreme precision for even the most sloppy alignment.
Most costly method, too, I bet.
You ought to know better.
Time for a redesign.

SM

 

[mitmeche]

Easy on the engineer

I know its hard but that is what the customer wants.

I generally, don't try to make things hard on the machinists but if it was easy everyone could do it. If its too hard its too expensive which is not what I want.

My problem is that my end customer wants the grid lines to span across two parts that are threaded together. Isn't there some way to specify the "phase" of the threads so when they are threaded tight, the lines match? Oh, and I can't put the lines on after I thread them together.

 

[HuFlungDung]

You can specify a thread phase angle, but its still going to be difficult to produce it without trial and error.

Thread milling, IMO, would be the best method to try to optimize the phase angle of the thread start because the path of the thread mill could be defined, ie., a specific entry point of the helix into the face of the part. Ordinarily, no one likely pays attention to where the thread mill's teeth actually are located axially, but one could begin to pay attention if it were warranted It still might cost a dummy part or two to get it set up, because the thread clearance would affect the alignment of the witness marks.

 

[gwilson]

Many guns have been made with timed threads on the barrels and actions so the barrels would screw in with the sights on top dead center.

 

[wesg]

Well sure, but it was a lot easier when they didn't have these damned CNC machines ;-)

 

[CalG]

Differing thread pitches into the two mating parts with a "loose" stud that allowed adjustment would be a dodge.

CalG

 

[pepo]

If I understand your post correctly you are asking how to draw this,not how to make it,right? Having a little trouble visualing your assembly but possibly a callout specifying an angularity of the grids or marks with the two details assembled at a certain torque value.

 

[toadjammer]

As stated the details either have to be machined assembled after threading or trial and error to develop because of the thread clearance. Gun barrels are settup at the factory to a master. Alot of the rifles aren't timed and the ones that are and have sites have the sites put on after the barrel is mounted. Another way to time a barrel is to take off the shoulder enough to line it up consistent torqueing has alot to do with this. And higher end guns that have timed screws are left long marked and measured for length and the slot is machined in the correct orientation.

 

[J_R_Thiele]

I am predicting a long and interesting discussion on this one.

A few questions.

Is the same shop making both the parts?

Can they be matched pairs or must they all interchange?

Must the two parts be one peice? Any chance of a threaded insert which can be glued or secured in some other way?

 

[register]

If this is a low-quantity run, I'd definitely put the lines in once the parts are already machined. Use whatever loctite seems most appropriate (blue maybe) and go to town. You will also have to specify a torque due to the the elongation in the bolt once tightened.

With all of this, I am imagining shallow lines (0.01) cut into the surface of the two parts in a grid pattern. If the pattern is heavier, then the loctite may not be sufficient.

If this all makes no sense, say so and I'll give it another shot tomorrow, but I have a strengths of mat'ls test tomorrow to study for (so that I can actually calculate things like the elongation of the bolt!)

Good Luck,
Henry Wettersten

 

[10KPete]

I was chief engineer on an oil well logging tool project where, because of the way the electronics worked, we had to have fairly accurate timing of each section of the tool.

This meant that each part of each tool section had to be timed to it's mating part. ie: the head connector, pressure tube and lower connector had to align. All those parts were threaded together with 29* stub acme and o-rings for pressure seal. The head and lower connectors had index features (keys) in them so they would align.

Now, this hardware was low volume production stuff mind you.

The head and lower connectors were turned from solid 17-4 PH, 4" OD and mostly hollow. Lotta chips.

The pressure tube was 4" OD x 1/4 wall 4130 DOM and, depending upon the tool section, 4' to 10' long.

After the connectors were screwed into the tube, 'locking' was achieved with a disc of sstl 5/8" dia set in a counterbore in the connector and overlapping onto the end of the tube. Key held in with a flat head screw. This kept the assy from unwinding yet allowed techs easy access for servicing the guts.

A master gage was made with internal threads on one end and external threads on the other. The key feature was marked with a scribe line along the length of the gage.

One end of the tube would be finished, marked with the gage and the mark extended to the other end. Second end would be threaded a bit deep then faced just enough to allow the gage mark to align.

The connector chunks were done CNC and when it came time to thread them the operator would align a mark with a mark on the chuck. The machine would then cut the thread in time with that mark.

We were only looking for a tenth degree per part (a couple tenths degree would have been fine) and the process seemed to work OK.

Long winded but I hope it makes some sense.

Pete

 

[John in CA]

I am probably missing something totally obvious here, so forgive me, but could the one part be screwed into the other and then the holes with the threads that need to line up timed be drilled and tapped into the assembly in one shot?

On edit, I think I understand the problem a little better after re-reading. One part must screw into the other and a grid pattern of features must line up when the threads bottom out. My question now is, a grid pattern of what? Holes? Bosses? V-grooves? If it's holes, then my original question stands; could the two parts be assembled and then the holes drilled?

 

[DaveE907]

Cameron, you don't provide enough information to adequately describe your problem. Are you talking about assembly of any two parts drawn from stock assembling as you require or are you talking about two parts meant for each other which can be dry fit before applying your grid pattern?

"I want" is an alarm to most machinists and chief engineers. I'm guilty of being both and I mean that comment kindly.

This is not a machine drawing 101 documentation problem. Perhaps notes and specifications are in order. Or maybe consideration of a cost effective design change is warranted. We don't know.

Pete detailed a successful solution for a real problem. (Really enjoyed it, thanks Pete)

If design isn't working without the fluid consultation of machine shop, production, all the people involved in making this whatsit, problems are ensured.

 

[Carl Darnell]

Yea I'm confused too.

Quote, I need to align a grid pattern across two threaded parts. One part screws into the other and the resulting pattern must line up. This issue comes up rather frequently and I don't know the proper way to specify the thread on each part so my machine shop will do the right thing and the parts will come out like I want.

When I first read it I thought he didn't know how to show a thread on a part. Aparently thats not so or maybe it is. I have a problem with, "one part screws into the other and the resulting pattern must line up".

Just what is "the resulting pattern" also called a grid pattern.

Cameron, can you explain this in common terms rather than your engineering jargon? Perhaps a drawing?

 

[mitmeche]

I am predicting a long and interesting discussion on this one.

A few questions.

Is the same shop making both the parts?

Can they be matched pairs or must they all interchange?

Must the two parts be one peice? Any chance of a threaded insert which can be glued or secured in some other way?

yep, same shop makes both parts.

matching parts ok.

I like the threaded insert idea.

 

[David Utidjian]

Without further details the way I am envisioning this part or assembly is:

A right circular cylinder of some diameter.
The cylinder is cut into two cylinders at right angles to the long axis.
One of the cylinders is drilled and tapped on one end.
The other cylinder is turned and threaded on one end.
Both cylinders can now be screwed together to form a single cylinder, same as the original stock but a bit shorter.

If the assembled cylinder is carefully made and screwed tightly together it can be turned from end to end on the lathe. This will make the seam between the two halves practically invisible.

With what appears to be a single cylinder... scribe, etch, engrave a grid pattern over the surface of the cylinder.

Since the grid pattern is scribed onto the assembly while assembled you are insured that it 'lines up'. BUT if you disassemble the two halves of the cylinder and then screw them back together again the odds of the grid patter lining up perfectly again are not very good. Take it apart and put it back together again multiple times and the alignment of the pattern will drift considerably. This is due to all sort of factors... wear in the threads, temperature, wear in the seating surfaces, variations in torque to assemble... even if the torque is meticulously kept constant it is still unlikely that the cylinders will ever re-align their grid patterns perfectly.

Yet, in a way, we do this all the time when we zero a micrometer. A micrometer with clean carbide faces will return to zero ( the zero tick on the thimble aligns with the 1/10" line on the barrel ) consistently for years. Even when frequently used. The repeatability in the zero setting of a micrometer is achieved by having a consistent torque limiter (ratchet or friction thimble, 'feel') and fine precision ground threads that are kept clean, and finely finished, very hard, and easy to clean mating surfaces (the anvils) and more-or-less consistent temperatures when zeroing.

If the "grid" in a micrometer becomes mis-aligned we can adjust it. Usually by turning the barrel slightly with a special wrench or adjusting the the fixed anvil, or adjusting the relationship between the thimble and the threaded rod slightly. On a digital mic just press the zero button.

When micrometers are made I don't think they "time the threads" so that marks on the thimble align with the marks on the barrel. I would guess that most micrometer thimbles are not interchangeable.

Well anyhow... just trying to think through the problem.

-DU-

 

[Perk]

As previously mentioned the features that require radial alignment should be machined after the threads generated, if possible. If not, parts can be fixtured to obtain alignment provided they are machined on a CNC with cycle to cycle spindle/axis orientation capabilities.

As for a callout for this condition:

1> NOTED FEATURES TO BE (RADIAL LOCATED xxxxx or IN ALIGNEMENT) WHEN NOTED PARTS ARE MATED AND TORQUED TO xxx (FOOT / INCH) LBS.