How to hold this part...

[TorontoBuilder]

I want to make a bunch of collets from 1/8" to 5/8" that have a 20mm wide waist section whose diameter varies based on the bore diameter that is 2mm wider than the bore to leave side walls of 1mm thickness.


Small capacity collets have a very narrow waist.

while the largest (5/8") has almost no waist.

The widest diameter at the end of the taper is 22mm, the taper is 20mm long at 5 degree angle. The straight shank is 18mm diameter by 30mm long.

I figured I'd start by turning the straight shank down to 18mm for a distance of 30mm. Then drilling hole 15mm deep with a #19 drill bit, and then tapping an M5 x .8 thread.

Next I'd extend the bar out about 35mm more and part off the blank to an overall length of 50mm. Once the straight shank has been turned on all of the blanks I'd switch to a 20mm collet for step two.

Facing off the 22mm end of all the blanks. Then I'd set the compound to turn the 5 degree angle.

So, where my issues come in are that I considered using 1045 which I'd heat and give an oil plunge to harden. I'd have to do this after I drill and ream each bore for the collet sizes, then turn down the waist dimension on each, and finally make the 3 slit cuts. Because of potential distortion I'd thought to make the collet marginally over sized and grind to the final dimension.

I can attach a 20mm stub to each collet with a 5mm screw to hold the collet in the 5c collet punch former to grind the collets. I'd have to insert a end mill into each collet to keep the bone from collapsing while grinding the taper, and I'd have to make the stub at the beginning of the process, and make a slight recess in the 5mm bore of each collet for the stub to mate with. Then mount each blank on the stub during to drill the collet dimensional bore to ensure it is concentric with the stub mount.

During the grind the straight shank and the taper will be ground into concentricity with the stub mount so the overall features should all be concentric.

Anyone see potential issues with this, or a better way to do any of the processes?

 

[TorontoBuilder]

The stub...

18mm diameter, 10mm wide x 3mm tall projection, and 8mm threaded hole thru.

Stub mounted to the straight shank to turn the taper to rough dimension and to bore and ream the collet nominal diameter.

The reason the stub is threaded for 8mm screw is to be able to jack the collet off the stub if it gets stuck

The previous process would need to be modified to bore a 10mm diameter hole 3mm deep in the collet straight shank prior to drilling and tapping for M5 screw.

 

[Perry]

My amateur thoughts. The total length of each collect will be 50mm. The stub that will be holding the collet you are working on will be 10mm? This stub will be held in a 5C collet?

You might have too much overhang on your work piece

Maybe leave a sacrifical stub on the collet until you have completed all the operations. Then cut/grind off the stub as the final step.

.

 

[PeterT]

I want to make a bunch of collets from 1/8" to 5/8" that have a 20mm wide waist section whose diameter varies based on the bore diameter that is 2mm wider than the bore to leave side walls of 1mm thickness.
Small capacity collets have a very narrow waist. while the largest (5/8") has almost no waist.

That sounds like a ton of work. But I'm not really clear on the the design intent. Maybe you can elaborate on how it will be used or what it will accomplish. I assume by waist you mean the annular thickness between the bore & the relief section OD? Using 5C collet as example, they are counterbored from the rear regardless of size so most of the material is removed. Only a portion of the nose segment, the holding area, is 'thick' for small bore sizes. I think that design serves to both accommodate the threads & makes spring collapse pressure somewhat equal across the range of holding sizes. You probably know this but typically they also have 2 precision ground surfaces - the tapered nose and the cylindrical guiding portion to the rear. For this, a single collet body is preferred for concentricity over a separate stub end (if that is removable?). Then of course some kind of inner/outer/both thread to draw it in & release. Anyways, I'm sure you know how a collet works, I'm just kind of confused by what you are proposing.

 

[Susquatch]

Silly question @TorontoBuilder.

Your design looks an awful lot like an R8 collet. Instead of making all new collets (which is a crap ton of work), why not change your overall system to use R8 collets instead. Or perhaps some other existing collet system.

 

[1018Machine]

the relief section OD? Using 5C collet as example, they are counterbored from the rear regardless of size so most of the material is removed.

This ^

I also think you are adding unnecessary work, which IMO will complicate things adding lots of potential for inaccuracies. I think back boring would be best.

What is the intended use for ? Something specific or general use?

 

[TorontoBuilder]

Silly question @TorontoBuilder.

Your design looks an awful lot like an R8 collet. Instead of making all new collets (which is a crap ton of work), why not change your overall system to use R8 collets instead. Or perhaps some other existing collet system.

These are for the Harold Hall's end mill sharpening accessory for his advanced grinding rest.

They're an odd size, and I had thought about instead adapting his spindle to take MT2 collets since I have some MT2 collets. I'd also thought about adapting his spindle to take ER20 collets since I have those as well.

I'd have to adapt the entire accessory to use 5C collets which is my preferred collet. I may just do that.

I only need this to sharpen the side of the flutes, since I'll use the other style end mill sharpener to sharpen on the surface grinder

 

[PeterT]

Regarding the rear threaded element of the proposed design, I was also going to reference an R8 collet for comparison purposes. When I look at some of the generic cross section pictures it kind of appears the inner section is further bored out to a diameter larger than the thread pilot hole. Then it either reduces or increases as a function of the grip diameter near the nose. I'm going to have to have a closer look at one in the shop. Funny the things you don't normally think about. All I can say is the few collet like gripping fixtures I've made with slits & whatnot seem to consume many (happy of course) hours in the shop & no heat treating or grinding was involved.

 

[PeterT]

We posted the same time. I figured it might be for some kind of fixture. I'm not to familiar with Hall's design but maybe an example is my tooling head for TCG. It uses an AliExpress ER16 shaft & ER collets of course. Precision ground on all surfaces & very reasonable cost IMO. The shaft counterbore is sufficently deep enough for my needs in this holding range, but that may vary in your requirements & may have to step up to bigger ER or bore out the clearance. I have not machined it so I don't know how hard it is. You can also get different (lower OD profile) nuts so they can get closer to the wheel.

 

[PeterT]

Another consideration for holding shanks like end mills. IMP endmills typically have nominal shanks but MET endmills are typically shank = cutting diameter. That's adding up to a lot of custom collets. This is exactly why I went with ER system replacing the (POS) R8 system in my Accusize TCG. It may not fit all applications but making the tooling head is less work & potentially more functional than making collets. There may be perfectly good reasons for Hall's design, or maybe it preceded what are now considered inexpensive decent quality collets/holders from China.

 

[TorontoBuilder]

We posted the same time. I figured it might be for some kind of fixture. I'm not to familiar with Hall's design but maybe an example is my tooling head for TCG. It uses an AliExpress ER16 shaft & ER collets of course. Precision ground on all surfaces & very reasonable cost IMO. The shaft counterbore is sufficently deep enough for my needs in this holding range, but that may vary in your requirements & may have to step up to bigger ER or bore out the clearance. I have not machined it so I don't know how hard it is. You can also get different (lower OD profile) nuts so they can get closer to the wheel.

View attachment 55615View attachment 55614View attachment 55616

I think I'll adapt this to my purposes. But I'll find this with ER20 nose. Thanks

 

[Susquatch]

Another consideration for holding shanks like end mills. IMP endmills typically have nominal shanks but MET endmills are typically shank = cutting diameter. That's adding up to a lot of custom collets.

Yup, yet another reason why Imperial is a better system than Metric......

 

[gerritv]

Shank diameter and measurement system benefits aren't really related though are they?
In my purchasing of metric end mills (as in affordable carbide) from AliExpress the shanks are 6, 8, 10 or 12mm. The largest I use is 8mm as the price goes up rapidly. Typical ranges cutting 1-6mm are 6mm shank, 7-8 are 8mm shank, etc. My R8 collets most used for milling, 6 and 8mm, 3/8. I avoid all other cutters since IMO there is no benefit to wierd sizes in hobby work. Depending on the store you might see 4mm shanks as well.
Engraving cutters and pcb mills/drills are usually .125" or 3mm.

Much like the .25->.375->.5 in progression. Except of course in the imperial world there are those 5/16, 7/16 and other off sizes.

 

[PeterT]

Typical ranges cutting 1-6mm are 6mm shank...

Its a good point. They are offered both ways. (I shop in the same place). For some reason most of the ones I've purchased are straight shank

 

[Susquatch]

Shank diameter and measurement system benefits aren't really related though are they?

The main benefit of imperial is my sanity. I need fractions in my life. Who wants a metric chunk of pie. I want 1/4 and if I'm hungry a half. My wife will take 1/2 of a 1/4 though.

In my purchasing of metric end mills (as in affordable carbide) from AliExpress the shanks are 6, 8, 10 or 12mm. The largest I use is 8mm as the price goes up rapidly. Typical ranges cutting 1-6mm are 6mm shank, 7-8 are 8mm shank, etc. My R8 collets most used for milling, 6 and 8mm, 3/8. I avoid all other cutters since IMO there is no benefit to wierd sizes in hobby work. Depending on the store you might see 4mm shanks as well.
Engraving cutters and pcb mills/drills are usually .125" or 3mm.

I dunno. I could swear I've seen 4.5, 6.5, etc. Heck, I think I've seen 4.3.

Much like the .25->.375->.5 in progression. Except of course in the imperial world there are those 5/16, 7/16 and other off sizes.

Those imperial ones sure are beautiful aren't they!

Its a good point. They are offered both ways. (I shop in the same place). For some reason most of the ones I've purchased are straight shank

Ya, that makes sense Peter. Explains what I have seen too.

All kidding aside, I was born in an imperial world and despite being able to work in metric, still feel most comfy with imperial. Basically, I think and visualize things in Imperial. My kids and grandkids are all hopelessly metric. Sheesh......

 

[mbond]

At the risk of wandering off into the weeds, the imperial system does have some important advantages

The metric system was developed with a great emphasis on uniformity - everything is some kind of power of 10. Within a certain unit (meter, gram, etc.) there is an absolutely fixed scale, and when combining units together the same continues.

But every unit in the Imperial system was developed from an applicability or fitness for the task. And the very fact that the conversions and combinations are _not_ powers of 10, makes it easier to spot errors if you have a sense for what the numbers should be.

These days I spend a lot of time checking for errors rather than doing direct work

 

[Susquatch]

But every unit in the Imperial system was developed from an applicability or fitness for the task. And the very fact that the conversions and combinations are _not_ powers of 10, makes it easier to spot errors if you have a sense for what the numbers should be.

Never thought of it that way! Good insight.

Thanks for that.

 

[trlvn]

From the original George Washington skit on SNL:

"Minuteman: How many yards to a mile?

Washington: No one knows."

Craig

 

[TorontoBuilder]

From the original George Washington skit on SNL:

"Minuteman: How many yards to a mile?

Washington: No one knows."

Craig

this is so damn funny, until you start to bake large batches of christmas cookies and you have to ask yourself, is this a Canadian or imperial recipe, or did my cheap ass dollar store import cheap chinese measuring cups meant for the US market or real proper imperial measuring cup.

I'm lucky in that all heirloom family cookie recipes are american origin, but unlucky that I used to go to the US every christmas and had to try to figure out what my gas was really costing me per imperial gallon

 

[TorontoBuilder]

This thread got away from how to hold the part and the planning for all the work sequences I wanted to solve, but that is okay because it prompted me to rethink the whole idea of do I really need to make this part at all.

The reality is, work and tool holding is an area where there are a plethora of options. In this case I'd be better served to modify the end mill sharpener spindle to accept a different collet rather than use the custom collets of Harold's design, or to use a more modern end mill sharpener design, and/or buy a professional sharpening system.

I have a large investment in grinders CBN and diamond grinding wheels and no real budget so the professional system is out for now.

So I'm going to punt.

I'm going to try to modify one of my 5C collet spindexer to sharpen the side flutes on end mills and use this tool to sharpen the ends