Best process for long through internal keyway with strict positional tolerances?

gregormarwick · Aug 12, 2020

135mm long turned part, 16mm h6 centre relieved bore, 4.985-5.015mm internal keyway straight through.

Keyway position is specified as a feature of symmetry with the bore diameter as the datum, and a tolerance of 0.0127mm. Depth tolerance is a bit more open at +/- 0.1mm

I don't know the purpose of the part, but knowing the end user, I suspect it is for sliding torque transmission inside a linear actuator of some description.

This is beyond our in house broaching facilities, so it will be subbed.

Question is, given the tight geometric tolerance, who should I sub it to? Would this be difficult to wire? Broachable?

mneuro · Aug 12, 2020

I would have quoted it as a wire EDM part. No real difficulty going that route.

RJT · Aug 12, 2020

Wire EDM definitely. It's an everyday kind of job for someone who does accurate EDM work. Assuming this is very low volume work. If it's high volume, then broaching with the right kind of tooling and equipment to do it. You would need high volume to justify the cost of broaching, but it would be much faster.

Peter Neill · Aug 12, 2020

Wire EDM for sure, have you got someone local to sub it to? Happy to recommend someone down this way if you need it.
Edit: If there's a possibility it's a sliding member, is there a surface finish spec for the keyway? May be worth establishing this first if you go the wire EDM route, then talk to subbie about achieving it

gregormarwick · Aug 12, 2020

Thanks for the replies all.

Peter Neill said:
Wire EDM for sure, have you got someone local to sub it to? Happy to recommend someone down this way if you need it.
Edit: If there's a possibility it's a sliding member, is there a surface finish spec for the keyway? May be worth establishing this first if you go the wire EDM route, then talk to subbie about achieving it

We do have a goto for wire jobs, but if you have someone that you know to be good I'm always happy to have more options.

Re. surface finish; That's precisely why I curious about broaching, seems to me that the surface texture as wired is not really ideal for a sliding surface, however there is no specific surface texture called out on the keyway, just general of max. 3.2μm. Curiously, the h6 bores and certain other features of the part are electroless nickel plated post machining, but drawing states the keyway is specifically to be masked off, which would lead me to think that the key is actually static in this part, and sliding on the shaft instead. However, there is nothing in this part to retain the key, so....

Peter Neill · Aug 12, 2020

Sent you a pm with contact details

implmex · Aug 12, 2020

Good morning All:
While I agree in principle that wire EDM is an excellent option for this project, I want to be sure all understand its limitations too.
We tend to think of wire as the super accurate magic bullet that can just make your part to within tenths at the push of a button: that is actually only partly true.

The difficulty is, and always will be, aligning the part on the machine table and then picking up the location accurately so the feature gets put in the right spot.

So vertical alignment of the bore is critical, and there are alignment toys you can bolt on the machine to help you adjust it and clock it in, but you still have to get it right, and there is another problem.
The wire may not be perfectly aligned with the Z axis motion and therefore if you clock in the part using the Z axis you might still not get a cut that's perfectly vertical.
I have basically the same wire squaring device that everybody else uses, and when I use mine a few times in a row, I will often have a hard time getting the device to agree with itself two times in succession.
There are lots of reasons for this and I won't go into them, but it's an area of uncertainty that can drive you batty if you care to get it right.
There are ways to check it; the best being to actually cut a test part and then interrogate it; a contributor here years ago advocated cutting a dowel hole in a perfectly square block and then checking that before critical parts are cut.
I do that now too, and I've had my bacon saved a few times by that advice so thank you most kindly to whoever it was who first suggested it.

It can be similarly difficult to pick up the center of a bore reliably; as with wire squaring I can run a center finding routine three times and get three different results.
The obvious question is: which one was correct...the less obvious question is; how do I confirm the one I picked was the right one?

So the take home is ; if I skim the bore and then cut the keyway I can get to superb relationship and dimensional accuracy really easily, but if you just want me to cut the keyway to an existing bore and still want it in the proper spot within tenths, it's a much harder proposition.

Not that it cannot be done by a careful operator, but it's not quite the slam dunk it's often portrayed to be.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Scottl · Aug 12, 2020

Just a comment from the sidelines.

This thread is an example of the collective wisdom of PM and how quickly they can suggest solutions to real world problems.

PRACTICAL Machinist - very aptly named!

DDoug · Aug 12, 2020

The "good" thing of WEDM is you are basically "Single pointing" that keyway.

With a broach, it floats as it goes down the bore.

I was going to suggest a single tooth key cutting machine, as able to hold a tighter
tolerance, but think the WEDM is better.

1953chevB · Aug 12, 2020

hard to discuss with out reviewing a print,
on EDM it would help to have a ground bore that the EDM machine can register , same for internal broach that slip fits to the pilot diameter of the broach.
but also a ground or very close turned outside diameter for locating in a pot or some fixturing. or soft jaws that are very concentric.
the pilot diameter has to locate on the smallest minor diameter , send a quote to broach masters in California USA , or broach house that is good near you, since you the UK
and see if they will bite.
while EDM is great, it is not cost effective for high volume in my opinion but I have done it many times when there was no other choice if the customer allows it.
they get hyper about recast layer and post hydrogen embrittlement but I have not had an issue with either.
I specify .0005 max recast layer and remove it and post bake if it is higher than 40 HRc hardness. to remove embrittlement.

Milland · Aug 12, 2020

Implmex's cautions are very helpful - thanks for that, Imp.

If the extra wire time isn't an issue, an undersize slot can be cut, verified, then the final size cut after any needed adjustments.

Simple hand polishing with a fine sandpaper wrapped over an undersize piece of key stock would allow smoother movement if that's of interest. Wire the slot to minimum size, take the last few tenths out with the sanding (roughish wire-cut texture means it'll go fairly quickly).

One thing I like to do if I have the option is leave as large a corner radius as I can get away with in the inside corners to give a little help with the cracking risk. Maybe ask the customer about that?

Mike1974 · Aug 12, 2020

implmex said:
Good morning All:
While I agree in principle that wire EDM is an excellent option for this project, I want to be sure all understand its limitations too.
We tend to think of wire as the super accurate magic bullet that can just make your part to within tenths at the push of a button: that is actually only partly true.

The difficulty is, and always will be, aligning the part on the machine table and then picking up the location accurately so the feature gets put in the right spot.

So vertical alignment of the bore is critical, and there are alignment toys you can bolt on the machine to help you adjust it and clock it in, but you still have to get it right, and there is another problem.
The wire may not be perfectly aligned with the Z axis motion and therefore if you clock in the part using the Z axis you might still not get a cut that's perfectly vertical.
I have basically the same wire squaring device that everybody else uses, and when I use mine a few times in a row, I will often have a hard time getting the device to agree with itself two times in succession.
There are lots of reasons for this and I won't go into them, but it's an area of uncertainty that can drive you batty if you care to get it right.
There are ways to check it; the best being to actually cut a test part and then interrogate it; a contributor here years ago advocated cutting a dowel hole in a perfectly square block and then checking that before critical parts are cut.
I do that now too, and I've had my bacon saved a few times by that advice so thank you most kindly to whoever it was who first suggested it.

It can be similarly difficult to pick up the center of a bore reliably; as with wire squaring I can run a center finding routine three times and get three different results.
The obvious question is: which one was correct...the less obvious question is; how do I confirm the one I picked was the right one?

So the take home is ; if I skim the bore and then cut the keyway I can get to superb relationship and dimensional accuracy really easily, but if you just want me to cut the keyway to an existing bore and still want it in the proper spot within tenths, it's a much harder proposition.

Not that it cannot be done by a careful operator, but it's not quite the slam dunk it's often portrayed to be.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Not sure what brand of machine you have, but on older charmilles they had built in cycles to align the wire to the workpiece. I think you could even automate it by inserting the proper command into the .cmd file. Not 100% on wire alignment inside the .cmd, but I know you can do things like clearance, rotation, scaling, wire cut/thread etc.

Probably the biggest drawback to wire is when you start getting into taller parts you start getting bell mouth issues. I think most machines have compensation for this, but might take a couple parts to get it dialed in.

1953chevB · Aug 12, 2020

Mike1974 said:
Not sure what brand of machine you have, but on older charmilles they had built in cycles to align the wire to the workpiece. I think you could even automate it by inserting the proper command into the .cmd file. Not 100% on wire alignment inside the .cmd, but I know you can do things like clearance, rotation, scaling, wire cut/thread etc.

Probably the biggest drawback to wire is when you start getting into taller parts you start getting bell mouth issues. I think most machines have compensation for this, but might take a couple parts to get it dialed in.

I worked with suppliers (that said they can just tough off with the wire to pick up the centerline of the parts with ground surfaces,
never had an issue with tapering but it was depending on the wire diameter, any comments on this?
I always got good parts

Mike1974 · Aug 12, 2020

1953chevB said:
I worked with suppliers (that said they can just tough off with the wire to pick up the centerline of the parts with ground surfaces,
never had an issue with tapering but it was depending on the wire diameter, any comments on this?
I always got good parts

Yes, on a wire edm you are using the wire to pick up your parts, if that is what you are asking...?

The charmilles I ran had multiple pick up strategies, including, wire alignment to part (before picking up and edge or whatnot), external center, internal center, corner, edge, etc. I imagine taper would be dependant on wire size. If you were running .012" wire I would *think* it would "deflect" less than .008" wire. A couple easy thins to mitigate it from the start, increase wire speed and brake tension. I don't remember the name of the parameter you changed to adjust bellmouthing, but if you lowered/raised (not sure after all these years) the value it would basically burn hotter/cooler in the center, I'm actually not sure how it really worked, but you get the idea.

1953chevB · Aug 12, 2020

Mike1974 said:
Yes, on a wire edm you are using the wire to pick up your parts, if that is what you are asking...?

The charmilles I ran had multiple pick up strategies, including, wire alignment to part (before picking up and edge or whatnot), external center, internal center, corner, edge, etc. I imagine taper would be dependant on wire size. If you were running .012" wire I would *think* it would "deflect" less than .008" wire. A couple easy thins to mitigate it from the start, increase wire speed and brake tension. I don't remember the name of the parameter you changed to adjust bellmouthing, but if you lowered/raised (not sure after all these years) the value it would basically burn hotter/cooler in the center, I'm actually not sure how it really worked, but you get the idea.

Thank you for the response, I appreciate it, I had them cut gears, disk gears, keyways, splines, w/I .0005 on the pitch diameters and tooth profiles or surfaces
it worked very well
again thank you

edit I supplied the DXF files of the gears, and splines

implmex · Aug 12, 2020

Hi Mike1974:
Yes I'm familiar with those cycles and for run of the mill parts they are a very good tool to have.
However as I'm sure you know, a wire touch to establish a location or orientation is dependent on many things.
First is the position of the feature in 3 D space...we've covered that.
Second is the geometry of the feature you're interrogating compared to what we're all pretending it is...so "round" holes that aren't round, "cylindrical" holes that aren't cylindrical, "flat" surfaces that aren't flat etc etc.
Third, of course is surface imperfections, burs, contaminants, roughness and on and on.
All of these contribute to uncertainty, and the wire squaring or orienting routines you refer to will all be limited by those realities.

What inevitably accompanies that uncertainty is errors that are hard to compensate for and strike at random.

For example...I just aligned a workpiece 5 minutes ago, and picked up a center bore that had been put in with a milling cutter and deburred with a fine abrasive wheel top and bottom.
It's been checked with a gauge pin and was clocked in on the mill for previous operations and is pretty round (within a tenth so far as I can tell with a tenths clock spun at only one Z axis height.

So far as I know, it's a pretty good hole, but I center found it three times and was off a maximum of 2 tenths in Y.
That's great for many things but if I'm trying for best possible accuracy it's a problem.

So I edge find it on the diagonal, and orthogonally and then I keep on doing it until I can get 2 successive readings within a tenth.
Then I call it good and cross my fingers and pray to the wire EDM Gods.
I have no better way than this, short of finding a way to measure the part while it's on the machine, and that gets hard when working to tenths.

That's our reality as you know; it will always be our reality so long as we have to wire touch to establish where we are.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

1953chevB · Aug 12, 2020

implmex said:
Hi Mike1974:
Yes I'm familiar with those cycles and for run of the mill parts they are a very good tool to have.
However as I'm sure you know, a wire touch to establish a location or orientation is dependent on many things.
First is the position of the feature in 3 D space...we've covered that.
Second is the geometry of the feature you're interrogating compared to what we're all pretending it is...so "round" holes that aren't round, "cylindrical" holes that aren't cylindrical, "flat" surfaces that aren't flat etc etc.
Third, of course is surface imperfections, burs, contaminants, roughness and on and on.
All of these contribute to uncertainty, and the wire squaring or orienting routines you refer to will all be limited by those realities.

What inevitably accompanies that uncertainty is errors that are hard to compensate for and strike at random.

For example...I just aligned a workpiece 5 minutes ago, and picked up a center bore that had been put in with a milling cutter and deburred with a fine abrasive wheel top and bottom.
It's been checked with a gauge pin and was clocked in on the mill for previous operations and is pretty round (within a tenth so far as I can tell with a tenths clock spun at only one Z axis height.

So far as I know, it's a pretty good hole, but I center found it three times and was off a maximum of 2 tenths in Y.
That's great for many things but if I'm trying for best possible accuracy it's a problem.

So I edge find it on the diagonal, and orthogonally and then I keep on doing it until I can get 2 successive readings within a tenth.
Then I call it good and cross my fingers and pray to the wire EDM Gods.
I have no better way than this, short of finding a way to measure the part while it's on the machine, and that gets hard when working to tenths.

That's our reality as you know; it will always be our reality so long as we have to wire touch to establish where we are.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Marcus
all good and very good points
I guess sometime I don't explain facts, I know from experiences for EDM just like gear (sometime)shaping, (sometimes)hobbing, and more so for gear grinding holding very tight tolerance I was taught early on the blanks have to be flat, square and concentric. My jobs are normally ground w/I .0005 or less
depending on the tolerance of the part requirements.
so yes I am very much agreeing with you, I take it for granted and don't explain correctly. brain not firing correctly

JS · Aug 12, 2020

I don't think, that I've had any real problems, when cutting a key way using a CNC Wire EDM.

Milland · Aug 12, 2020

Stupid questions - why not rough the part, drill a ~14mm starter hole, then finish the bore and key in the WEDM? Then throw it on an arbor, and finish the rest of the geometry. As long as the wire isn't grossly off, all the feature will be in the correct relative positions and geometries.

Sounds simple to me, so I must be missing something...

implmex · Aug 12, 2020

Hi JS:
Me not either...but if it's a keyway that needs to be on location within tenths, its a bit trickier than it looks...that's all I'm saying.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Best process for long through internal keyway with strict positional tolerances?

Diamond

Cast Iron

Titanium

Stainless

Diamond

Stainless

Diamond

Diamond

Diamond

Cast Iron

Diamond

Diamond

Cast Iron

Diamond

Cast Iron

Diamond

Cast Iron

Stainless

Diamond

Diamond

Similar threads