Good dead length collet setup for a Haas TRT100

[implmex]

Hi All:
I've been working to commission a Haas DT2 with a TRT 100 on it.
I am looking to find the best dead length collet solution for it.

Those who know the TRT100 know it's TINY... 4" diameter platter.
The through hole is also tiny...maybe 1/2 inch (I haven't measured it.)
My choices for a dead length setup as I see it are:
1) A good quality baby 3 jaw chuck I can bolt to an adapter plate and then bolt it to the platter.
2) build a collet system with cylindrical backs and a tapered nose with a taper nose collet nut.
3) Try to find something...maybe a wire EDM self centering chuck like this:


This is 5 thousand bucks USD...kind of overkill but it is more accurate than a collet and it IS dead length.

Anyone have any other ideas?

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[Donkey Hotey]

Depending on the diameter of the part and how aggressive you want to cut, I'd be careful. I had a 2" round bar yoinked out of a 8" 3-jaw doing some very moderate cutting. Really woke me up. Snapped a new Niagara end mill too. Changed to the 6-jaw and didn't have any further problems. I don't like milling round stuff in 3-jaws unless they're fat or even pie soft jaws.

On the EDM fixtures: aren't those high-precision but, low holding force?

 

[crossthread82]

Royal makes a dead length mountable 5c collet block that is actuated from the side.

https://royalproducts.com/product-l...l-5c-16c-and-3j-key-operated-collet-fixtures/

 

[Donkey Hotey]

I just refreshed myself with the platter on the TRT100. I didn't remember if it had a zero point system incorporated or not. That's an interesting interface they have on the platter. If a collet closer won't do the job, I'd be tempted to roll my own solution.

I have a T5C, the TRT100's slow, awkward cousin. I've never had it in the machine (limited space around the machine and heavy). I'm currently designing the first thing I absolutely, positively need 3+2 to hold tolerance. It's getting installed in the next few days. I'll be very interested in the direction you go.

Relating the difference in our challenges: you have a flat platter and no way to really hold a round part. With the T5C, I have either the 5C to hold things (not dead length) or the Hardinge threaded spindle nose (2-3/16 x 10). I have a few chucks and faceplates to choose from but stackup from the axis of rotation gets big. You're in the opposite boat: platter close to centerline, tons of clearance but, nothing to grab round bits.

One of these:


The thing I don't like about the dead length collet closers is the floating collar that pushes up around the neck of the collet. It's all a trade-off.

 

[kustomizer]

These are good, I made an adaptor for one of mine to put it in a 5c indexer

 

[implmex]

Hi again Donkey Hotey, crossthread82 and kustomizer:
I am confronting this workholding problem because I have parts
that require a set of operations at each end with a flip and re-chucking on which I need to hold tight tolerances for length as well as good concentricity between the first feature set and the second.
The hard part is the length.
The setup so far (it's pretty hokey I confess) has been to modify a little 4 jaw chuck so it can bolt directly to the TRT platter.
Into that goes a shorty straight shank ER20 collet chuck with a stop threaded into the shank.

The fantasy was, that since the stop is dead length (mounted to the chuck body instead of the collet like a 5C stop) the collet would pull the part down onto the fixed stop and all would be lovely.

Well as it turns out it's still exquisitely sensitive to the diameter of the parts after the first ops, and how hard you reef on the collet nut.
I cannot reliably control the overall part length better than about +/- 0.002" from nominal.

The workaround was to slice the part from the stock with a saw in OP1 so I could control the length and then probe every last damned part for length for OP2.
What a PITA.

Ideally I'd find a nice, accurate, low profile, tiny 3 jaw or 6 jaw scroll chuck that can be bolted to the platter with a stop mounted to the chuck body.

It doesn't need to hold a lot...the most common parts are PEEK or aluminum and are typically 6 mm diameter or less and maybe up to 15 mm long, so all small stuff.
I have a similar setup (but bigger) for my Minimill and it works very well for length control...I can hold +/- 0.0005" with care, but my concentricity is not quite as good with a scroll chuck as it is with a collet.
I can live with that.

But my Minimill setup is way too big for the TRT100.

Crossthread82, that link you posted to the Royal website is really worth looking at.
I fear the whole works may be too big and take the top of the job too far from the A axis center of rotation, but I'll make up a mockup in Solidworks and see how it looks.

Kustomizer, that's a nice gadget too, and for my Minimill it could be pretty useful, but it doesn't look dead length, unless there's something I'm not seeing.

So I may try to knock up a quick and dirty custom collet that has a taper only on the nose with a matching taper on the nut, and a square shoulder to butt against on the collet body.
So unlike an ER20 or a 5C in that the collet doesn't move axially...it's held by the shoulder.
It'll be interesting to see if it stays concentric...I have misgivings.

If it works better than I expect it to, I'll build a proper hardened version.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[Freedommachine]

Marcus,

Would it be possible to pick up an exposed feature on the part with a probe and reset Z? (Assuming your machine has probing capabilities)

If it is only off in length but all other directions are repeatable, probing could be an easier way to work around it.

 

[implmex]

Hi Freedommachine:
Yeah, that turned out to be my workaround.
I established my blank length by milling all the features on one side and then running a stiff slitting saw as the last operation, to slice them all off to the same length.
That way when I flipped them I could probe the ends for height and then mill in the features on the second side.
But it was a pain in the ass to probe every part...I was hoping to just use a stop, but no joy.

You may well be right, and I will have to bite the bitter pill and just accept that if I want them accurate I have no choice.

BTW, the parts are so small that I can't probe for concentricity...the part sticks out of the collet only by about 1.5 mm and the probe ball is 4mm diameter.

So I probe a pin during setup and then I pray a lot.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[kustomizer]

The collet is threaded into the base and locked with a set screw in the collet slot, the sleeve on the outside of the collet moves in and out with air pressure, when I get to work I'll make some more pics

I have a couple of microcentric air chucks, NOS that would work for a job like this as I understand it.

• .00005" (0.0012mm) TIR is standard

 

[Donkey Hotey]

Small hijack here: is the DT-1 for the same people that bought the CL-1? Could they have bought any more extraordinarily wrong machines for the job and the money? Where's the giant godzilla facepalm meme? Maybe they could have surprised you with a VS-3 to do all those micro parts? I'm dying here. They can marvel over the 1-second DT-1 tool changes when they make 30 parts.

They needed an ST-10, a VF-2 and maybe a TRT-160 or 210. They'd have had change for you to buy all the tooling you'd have ever dreamed. Instead they saddled you with these oddball machines and expect you to hold PDOOMA tolerances? My sympathies--genuinely. Someone is patting themselves on the back for the great choices they made.

 

[SeymourDumore]

So unlike an ER20 or a 5C in that the collet doesn't move axially...it's held by the shoulder.

Marcus

If Hardinge's website didn't suck so bad, I could link to the proper item you need.
Otherwise, please look at this collet, and see if you can find the rest of the parts that go with it.
Rest assured, this thing WILL repeat in length within a tenth or two: https://shop.hardinge.com/All-Products/Workholding/Collets/Lathe-Collets/5C-Dead-Length®-Round-Collet/p/11590017000000


Found it!

5C Dead-Length® Collet Assembly less Inner Collet

<p><strong>Hardinge® Dead-Length Collet Assembly</strong></p> <p>Shoulders and faces can be machined to exact length, regardless of OD variations with the Hardinge Dead-Length Collet Assembly. Parts are loaded against an adjustable solid stop which is threaded into the inner collet. The inner...

shop.hardinge.com

 

[Donkey Hotey]

Found it!

5C Dead-Length® Collet Assembly less Inner Collet

<p><strong>Hardinge® Dead-Length Collet Assembly</strong></p> <p>Shoulders and faces can be machined to exact length, regardless of OD variations with the Hardinge Dead-Length Collet Assembly. Parts are loaded against an adjustable solid stop which is threaded into the inner collet. The inner...

shop.hardinge.com

Damn, thank you for sharing that. With all the 5C stuff I've seen, I've never seen that.

This is why I love trade shows and try to linger in obvious booths where I think there's nothing to see: Kurt, Hardinge, etc. There's always something like that in plain sight: "Hey, what's that? I've never seen that before." Also to let them talk. They often have some cool new thing they're excited to share or design improvement that isn't obvious.

 

[EmGo]

Could they have bought any more extraordinarily wrong machines for the job and the money? Where's the giant godzilla facepalm meme?

I was just thinking that. Maybe the best thing Marcus could do would be to teach them to not make really stupid choices in the first place. Being smart about getting out of a hole you dug for yourself ... jeeze. Not the best way to do things

 

[Donkey Hotey]

As mentioned earlier, my interest in this thread is due to currently designing a holding fixture for a 3+2 part I'm going to be making in quantities. Mine is going on a HA5C.

Based on the Hardinge collet @SeymourDumore shared: I'm thinking my own answer might be a screw-on fixture that fits the threaded nose of the rotary. Have the basic part shank diameter, a pocket and some pin-locations to pick up clocking.

Behind that, I'd bury a soft 5C collet, with the dome faced flat so it will fit beneath the screw-on fixture. With the two, the 5C collet would pull the part down but, the fixture on the nose would be the dead-length stop. It's held axially by the collet and held laterally by the fixture. The fixture gives something to indicate into position for the rotary zeroing.

For Marcus: some of the 5C collet closers also offer the threaded nose and could be tooled the same way. Don't know if you need the Z-stop to specifically be the back surface of the part or if you could datum from something further up. Or maybe leave extra material during manufacturing and deck that off in a second operation?

 

[memphisjed]

Vice with customized jaws? Less moving parts, less is more.