TTS ER-20's are killing my collets! Anyone see this before? Know what's happening?

[kstrauss]

Perhaps people do use collet chucks for drill holding....horses for courses etc........for the smaller drill size, a 1 Morse taper drill shank can't be beat......no moving parts and they don't run out.....ever......that's an industry standard.
Ian.

I should probably keep quiet but...

I've never seen MT shank bits smaller than 1/8-inch or so. My drills are mostly #43 or smaller.
I've never seen an ATC for MT shank tooling. Certainly Tormach doesn't offer such.
I am not aware of a professional CNC mill -- Haas, Okuma, Mori Seiki, etc -- with a MT1 spindle

[handlewanker]

Hi....that is as maybe.....I have a couple if Morse tool holders with ISO 30 shanks made by Emco with pull pins too, so they were intended for CNC work.

Not everyone has or wants an ATC....it would be a fun thing to have and let the mill do the complete job from start to finish untouched by human hand etc.

The point of Morse tooling is it's simple and very retentive....no moving parts etc and has a very slim body compared to a drill chuck.........small drills can be held in 0 - 6mm keyless drill chuck and also held in Morse tool holders as required.......it's just one way to get a tool to rotate in a spindle.

For a comparison, the ISO 30 tool holder I mentioned has a 40mm body diam outside the shank with the #2 Morse taper inside....there is a 50 mm long part that extends from the tool holder and contains the Morse taper.....any Morse taper shank drill taper held in the taper is flush with the end of the holder......just the drill itself is outside.

BTW....the actual spindle of the machine is not Morse taper it can be any taper like ISO 30, R8 etc....the tool holder that fits into the spindle has the Morse taper.

And for the record....Morse taper goes down to 0 Morse for small drills too.
Ian.

[mactec54]

Hi....to clarify my statement of the single slot ER collet proposal.....the Autolock S type collet has only 3 slots but is dedicated to hold a specific size cutter shank diam.

It's also only suitable for threaded shank cutters as the cutter torque tightens the tool with loading......and you don't need 50+ foot lbs of torque to tighten the nut.... the cutters are inserted finger tight.

I've made many split bushes to hold odd diams in a larger collet in my old turret lathe when I had to hold an odd size material diam, and with a single split the bush held the material securely.

The same method would work well for the ER type chuck and solid collets with just one split.
Ian.

The same method does work and they already have Tap collets for the ER system just like that with 4 slots, your Autolock s type chucks are no where near as good as a ER system there collets only grip the top of the cutter shank 3 point contact, and you can not use them hand tight, you use hand tight then screw the cutter down to the center point, back the cutter off a few thousandth , and then lock the nut

[handlewanker]

Mac....with the S type Autolock you do not back the cutter off at all....that is stated in the instructions for the S type........the OLDER type with the collet and sleeve closer does need the nut backed off before you screw the cutter in and onto the centre point......when the nut is tightened the collet and sleeve close on the cutter and force it down onto the centre point, but by doing so the nut is actually clear of the body face.

Anyway, the torque of the cutting force in the old type will rotate the cutter slightly and as it's screwed into the collet push it up and so exert more clamping force.

In case I have been confused with the previous explanation, my work experience with the S type was that you mount the collet in the closer, screw the closer with collet into the body until it touches the body face, then screw in the cutter onto the centre point and then tighten the nut.....the instructions specifically stated the old way of mounting the cutter was not to be used with the S type.

BTW.....the collet does make full body contact for the full shank length as it's internally ground round to a specific diam for the cutter for it's length and as the collet is split down to the bottom end, when it's tightened with the closer or nut, it's a total body length grip not just at the top; which otherwise would have to have the cutter a loose fit in the collet to allow the collet to close and make contact only at the top.

We always inserted our cutters in the collets for the S type by holding them with a piece of rag, due to the close fit that the cutter has with the collet even in the relaxed state....this was to prevent slicing your fingers on the cutter sides when you grip it.........at no time would you get only a 3 point contact with the S type system.
Ian.

[mactec54]

Mac....with the S type Autolock you do not back the cutter off at all....that is stated in the instructions for the S type........the OLDER type with the collet and sleeve closer does need the nut backed off before you screw the cutter in and onto the centre point......when the nut is tightened the collet and sleeve close on the cutter and force it down onto the centre point, but by doing so the nut is actually clear of the body face.

Anyway, the torque of the cutting force in the old type will rotate the cutter slightly and as it's screwed into the collet push it up and so exert more clamping force.

In case I have been confused with the previous explanation, my work experience with the S type was that you mount the collet in the closer, screw the closer with collet into the body until it touches the body face, then screw in the cutter onto the centre point and then tighten the nut.....the instructions specifically stated the old way of mounting the cutter was not to be used with the S type.

BTW.....the collet does make full body contact for the full shank length as it's internally ground round to a specific diam for the cutter for it's length and as the collet is split down to the bottom end, when it's tightened with the closer or nut, it's a total body length grip not just at the top; which otherwise would have to have the cutter a loose fit in the collet to allow the collet to close and make contact only at the top.

If you had used them enough you would of seen when you removed a cutter that there was a compression mark around the top area of the cutter shank

We always inserted our cutters in the collets for the S type by holding them with a piece of rag, due to the close fit that the cutter has with the collet even in the relaxed state....this was to prevent slicing your fingers on the cutter sides when you grip it.........at no time would you get only a 3 point contact with the S type system.
Ian.

It appears you have the 2 types mixed up the first type that was manufactured in 1942 was the hand tight type photo below the S type is in the post above where you need the wrench to lock them up, both only have the top part of the collet clamping around the shank of the cutter just look at the photo

What you are saying is correct, until you break out the back of the cutter which happens even with the S type I had both and used them for many years, now they are just a talking piece, at the time there was not much else to use, as soon as ETM ER collet system hit the market just over 30 years ago, they have never been used since, as well as lots of milling cutters, for them, if anyone are puzzled with this, these auto-lock chuck system has to have cutters with the thread on the end for it to work

The collet and the inside of the Nut where the collet slides in is parallel, only the taper on the top of the collet does the compression, so only have the top of the collet clamping the cutter, 3 point contact, the cutter fit in the collet and nut, are a close fit, this tooling today is a dinosaur, if you are trying to compare them to todays tooling, all you can say is they where good in there day

[handlewanker]

Hi Mac.......I don't remember ever seeing an Autolock chuck with a design like you posted in the last post.

The early type I mentioned had the internally threaded collet and a sleeve with two pegs that fitted closely in the chuck body, relying for concentricity on the centre point and the top sleeve taper.

The early type had the sleeve with two pegs that engaged in two small holes on either side of the centre at the bottom of the chuck bore to prevent the sleeve from turning when tightening the nut.

The S type has the threaded collet alone, no compressing sleeve, with a distinctive T shaped collet end that engaged in a slot in the bottom end of the nut.....both types had the nut screwing into the chuck body but flush against the chuck body face.......this is totally different to the design you posted last.

I still have the S type with an ISO 30 shank ( and a swag of threaded cutters from way back when) that I might still use for a jig borer 2 morse spindle conversion I have in mind.

The Autolock chuck was an industry super standard for many years.....pity it needed the screwed shank cutters........when I came in contact with the Bridgeport R8....that was the answer to all prayers.

BTW.....in the past, plain shank cutters were called Bridgeport type because they didn't need the thread.
Ian.

[philbur]

I have to wonder why an ER collet has so many segments when a solid one with just one slot would work better, and you wouldn't have to pre mount the collet in the nut to fit it in the chuck.
Ian.

One important function of the nut is to extract the collet when the nut is slacken off. If the collet is not preinstalled in the nut this a wouldn't happen. Have you never felt the resistance as you hand unscrew the nut after you have loosened it with the wrench?

Phil

[handlewanker]

Hi.....yes I realise that factor........I have 2 chucks, one an ISO 30 and the other an ISO 40 shank size.....they are not ER 32 even though they look about the same dimension, the collet is different,,,,a bit longer and slightly bigger in diam.......I'm told they are Persque collet chucks......never seen that one around........as I only have one collet that fits both chucks and no source for others, I'll be selling them off on EBAY.

I got them with a packet of bits and pieces with the Ajax mill I bought........long story short......the nuts on both have ball bearing rings in the top section.....the retainer and extractor part is a spring ring that the collet slips into when you push it into the nut ....same effect as the ER type with the off centre top taper.

If you are familiar with the Fenner taper lock bush system for mounting pulleys to shafts you will see that what I proposed with the ER single split collet follows the same design principle as the taper lock bush which has only one split to allow it to compress and grip the shaft........the extraction method with a groove around the top end of the collet would remain the same with a spring ring that automatically expanded to snap into the collet groove.

The ER nut has a solid extractor ground slightly off centre into the top of the nut and requires the collet to be pre-mounted with a sideways movement.......if you don't pre-mount the collet it won't seat properly if at all.

My opinion still is that an ER collet with a single slot would be simpler to make even though it only fits one specific size and is not able to be squeezed down 1 mm like the ER system.

To go a step further, if the collet had a thread in the bottom end and the cutters were threaded too, like in the old days, cutter pull down could not happen......cutters with a thread would only cost a few cents more.
Ian.

[kstrauss]

I don't believe that tool pullout from a properly tightened ER collet is an issue. It has certainly never happened to me.

A collet with a single slot would have a very limited grip range. Perhaps more importantly it would require a huge force from the nut to cause it to grip the tool. Consider the difficulty of gripping a 1/8-inch bit in a mostly solid ER32 sized collet. Also, with an essentially rigid collet I'm unclear how your proposed extractor ring would allow the nut to be separated from the collet. Not bad ER collets are available for less than $10 so making them more cheaply would be of limited benefit relative to the other costs of running a CNC mill.

[philbur]

To go a step further, if the collet had a thread in the bottom end and the cutters were threaded too, like in the old days, cutter pull down could not happen......cutters with a thread would only cost a few cents more.
Ian.

Where do you buy carbide cutters with threaded shanks?

Phil

[tmarks11]

My opinion still is that an ER collet with a single slot would be simpler to make even though it only fits one specific size and is not able to be squeezed down 1 mm like the ER system

Yes, sure, make something that takes at least 4 setups, and requires both mill and lathe to manufacture, which offers no advantages over an endmill holder that can be cut on a manual lathe in a single setup. Something that should be made out of spring steel if you do it right, which will require careful heat treatment and quenching.

[handlewanker]

For the record....I do not intend to make collets,,,,,,,I have made many collets with a single slot for various purposes.......that they are practical is comparative to the Fenner taper lock pulley system that has a single slot taper bush to grip the shaft......a specific shaft size only......and that is by screwing in two screws.

With a full diam the exact size of the cutter, huge force is not needed to clamp the cutter.

I merely mentioned the practicality of the design in passing......I also suscribe to the specific size collet for the cutter and not the squeezing down method as suggested by the sellers.

BTW.........what fool uses 1/8" cutters in an ER32 collet chuck????.....surely the smaller size ER chucks that go down to 1/8" or 3mm will be less expensive and also less prone to vibration at high speed.

If you compute the range of cutter sizes from 3mm up to 20mm and compare them to the range of ER chucks, you will probably find that for the range of approx. 6 cutter shank sizes you need at most 3 chucks and 2 collets for each.....2 more collets for each if you want to have Metric and Imperial collets.

BTW......I only mentioned screwed shank cutters and collets as a solution to cutter pull down......that many people seem to have......it won't happen so it's purely design hypothesis.

There is another ....design hypothesis.....that caters for cutter pull down and that is with the single slot collet..........a small pin that protrudes into the side of the collet bore to engage with a small indentation or flat on the cutter shank......with this design the pin, which is loose would push into the collet bore when the collet is inserted into the chuck and into the indentation........the indentation or flat can be after market ground in to any cutter be it HSS or carbide.......just a design hypothesis.

I would have to add, that anyone who is plagued by cuter pull down can benefit from such a design without resorting to high tech special collets etc.
Ian.

[handlewanker]

Yes, sure, make something that takes at least 4 setups, and requires both mill and lathe to manufacture, which offers no advantages over an endmill holder that can be cut on a manual lathe in a single setup. Something that should be made out of spring steel if you do it right, which will require careful heat treatment and quenching.

I fail to see your reasoning........an ER collet has to be slotted from both ends with multipole slots.......commercial manufacture accomplishes this in mass production anyway......I suggested only a single slot.
Ian.

[kstrauss]

Just for clarity, this fool doesn't use ER32 holders for 1/8-inch cutters. I have only one ER32 TTS holder and only a 20mm ER32 collet.

Again, I have never had tool pullout from a properly tightened ER collet nor have I read of any such problems. Mentioning "design hypothesis" solutions to tool pullout is interesting but irrelevant to solving real problems. Using ER collets for drills offers numerous benefits including no drill pullout, lower cost than quality drill chucks, improved concentricity, lighter weights in an ATC and reduced Z-axis consumption. If your ER collets are used only for cutters than very few are required to accommodate all cutters of both imperial and metric diameters.

My FEA skills are insufficient to do the calculation but to easily insert a cutter into a single slot collet requires perhaps 0.0005 of clearance and even that small amount of collapse requires significant force.

[philbur]

I would have to add, that anyone who is plagued by cuter pull down can benefit from such a design without resorting to high tech special collets etc.
Ian.

A Weldon type tool holder achieves the same with much less complication, it doesn't get any more low tech!

Phil

[handlewanker]

Just for clarity, this fool doesn't use ER32 holders for 1/8-inch cutters. I have only one ER32 TTS holder and only a 20mm ER32 collet.

Again, I have never had tool pullout from a properly tightened ER collet nor have I read of any such problems. Mentioning "design hypothesis" solutions to tool pullout is interesting but irrelevant to solving real problems. Using ER collets for drills offers numerous benefits including no drill pullout, lower cost than quality drill chucks, improved concentricity, lighter weights in an ATC and reduced Z-axis consumption. If your ER collets are used only for cutters than very few are required to accommodate all cutters of both imperial and metric diameters.

My FEA skills are insufficient to do the calculation but to easily insert a cutter into a single slot collet requires perhaps 0.0005 of clearance and even that small amount of collapse requires significant force.

Well if you're talking about force to clamp an ER collet, the poster of this thread stated that he uses 50 ft lbs......on an ER20 collet chuck nut.........gasp, and I only use a 250mm long C type spanner by hand to tighten them......and if that ain't a significant amount to apply just for normal holding I don't know what is........he must be anticipating huge torque cutting forces.....on a 13mm max diam cutter?......... to warrant that much force........anyway, a single slot collet will grip if any force is applied to the taper.....at all.

With hindsight, I think I would agree that an ER 20 chuck would be quite neat and compact, probably more suitable and compact than an equivalent 13mm keyless chuck for holding the range of drills from 1mm up to 13mm,,,,.....given the squeezability factor of the collet and that the collets are not all that expensive either.

I might invest in an ER20 chuck or two.....they sell them with an ISO 20 shank size (to suit my mill spindle) on EBAY and I already have a collection of ER 20 collets for my water cooled spindle.

As it's getting completely off topic and irrelevant to the problem, I'll leave it as is, as it's just a hypothetical thought path.
Ian.

[handlewanker]

A Weldon type tool holder achieves the same with much less complication, it doesn't get any more low tech!

Phil

Weldon?.......is that the type that holds with a screw angled to one side against a tapered flat on the cutter?
Ian.

[RedneckPhysics]

Hi All,

A big thank you to everyone who responded to my original post! Plus, this has turned into an interesting discussion.

I wanted to get back with the results of a few test sooner, but things have been incredibly busy here (I'm a returning student and fabricator/researcher within the EE department here at NC State... it's fun at times, but I definitely don't recommend it!).
Amidst other issues with our 770's setup, I was able to test out some of the suggestions from GenInt and others.

First, I tested the compression angle of the tool holders themselves.. I did so by cleaning, deburring and degreasing the TH's and collets as best that I could, then manually compressing the collet into the toolholder by itself (no nut or tool), to see how the spacing along the length of the collet looked (front vs. back). Interesting results. Using a new 1/2" Tormach supplied collet (with LOTS of deburring to try to remove the mess of burrs left-over inside of the collet fingers), the collet would stay compressed easily in some TH's, and loosely/not at all in some (I imagine this would immediately indicate an angle variation). After fully compressing the collet, I found that it was actually so poorly machined, I couldn't get a definitive read on the segment gap spacing. However, it looked like both sides were compressing almost equally... if they weren't, it wasn't discernible with my eyes. I wanted to try with a 3/8", but I had no more available, and didn't really feel like doing this to the one high-quality collet I had left. Hopefully Tormach supplies new ones.... still haven't bugged them a second time about it, but it's next on the list.

Secondly, I coated the mating flats on one of the tool holders that didn't retain a collet and a brand new nut with bluing. I inserted an optical post (closest I have to a nicely machined 1/2 inch tool on hand), then tightened to LESS than the "death torque" (I believe I chose 40 lb. ft. this time). I'll have to try the 57 Lb. Ft. again if Tormach resupplies me. This time, the nut torqued-down with good, even flat separation, and upon removal, saw no noticeable galling. I got a contact pattern that was fairly uniform throughout. The larger diameter portion of the nut's taper (which galled previously) sure seemed to be getting more contact than the smaller diameter portion. Took pictures, but with the blued surface on an angle, you'd have an easier time photographing the loch ness monster.
I did the same again with one of the "dead", heavily worn collets and nuts (not certain if the TH was still paired with the original failure combination). Went up to death torque, but saw no deformation, so I went a little beyond (still nothing like before). Checked the Dykem pattern, and not really any different in the toolholder. The larger diameter portion of the nut that experienced galling before DEFINITELY got contact. All bluing was removed from this area.

So, while not conclusive, the tests seem to point towards a machining defect in the nuts, or a metallurgical defect as many have mentioned here previously.

I ordered some Moly/graphite based JetLube anti-seize for the nuts (side note, it's a low-toxicity chemical compared to the other high performance anti-seizes I researched... never hurts to accumulate less bad stuff in your shop). Hopefully that, and much less torque (except if I actually get some collet slip down the road) will help with preserving the rest of the ER combinations that I have on hand.

Since this is still an unsolved mystery, and I'm still curious as hell as to what happened exactly, I'm going to speak with a few of the metallurgists here in the Industrial Engineering department. I'm taking a metals AM course with a destructive test specialist, so I'll see if I can't ask him to help diagnose this, or at least use some of his lab's hardness test equipment.


Thanks Again,

Mike

[handlewanker]

Hi Mike.....I hope you aren't fooling yourself with the collet in the tool holder sans nut compression test.

By this I mean, when you insert a cutter in a collet of specific size to the cutter diam it becomes relatively a solid item......iE, it (the collet) can't be compressed further as the cutter will hold the back taper of the collet to that size........the nut will just clamp it in that position in the tool holder bore.

It's a different story if you use Imperial cutter diams in Metric collets or cutters 1mm smaller than the collet size, as you will have some movement by the collet back into the tool holder body as the collet collapses to grip the cutter......as much as a 1mm difference in cutter to collet size is stated by the sellers to be allowable.

I think that in this situation the collet segments could be twisted as they collapse, as they do not have any contact with the cutter diam until they are squeezed down.

There should be no discernible back movement of the collet in the tool holder with a cutter in it.....a few thou perhaps as the nut exerts a backward force against the collet, but if you mounted a dial indicator and measured the zero with the nut just touching the collet, then with the collet torqued up with the short spanner supplied for the purpose by the sellers......there should be very little difference.

BTW.....someone mentioned that high clamping pressure applied to the nut "can" cause the tool holder body to expand..........that's unbelievable.......but tool steel does have some spring in it when the wall thickness is thin......that would cause the collet to drive deeper into the tool holder for no real benefit.

One thing is for sure........I use ER32 tool holders on my manual mill and I would not dream of removing one every time I wanted to do a cutter change.....applying 57 foot LBs of torque with a tool holder still in the spindle is impossible...........I gasp to think of how much an ER32 "should" be torqued to.....I use the C spanner supplied to tighten the nut by hand with the spindle brake on.

BTW.....how do you get 57 foot lbs on an ER20 nut.......some ER20 nuts I've seen have two slots for a spanner..... due to the narrowness of the slots, huge torque forces would cause the spanner jaws to spring open

It still doesn't explain the galling that occurred, but I put that down to the case hardening skin being too thin and getting ground down in manufacture.....another story.
Ian.

[RedneckPhysics]

Hi Mike.....I hope you aren't fooling yourself with the collet in the tool holder sans nut compression test.

By this I mean, when you insert a cutter in a collet of specific size to the cutter diam it becomes relatively a solid item......iE, it (the collet) can't be compressed further as the cutter will hold the back taper of the collet to that size........the nut will just clamp it in that position in the tool holder bore......

Yes, I agree, the collet shouldn't have moved much at all with respect to the tool holder's taper. You bring up a great point, can't forget that there's always going to be some deformation going on with every part :-).

In the end though, not sure I really learned anything conclusive from the tests I performed.

As far as tightening the collet nuts to death torque: The ER-20 nuts Tormach supplied were the type with hexagonal flats. Definitely have to torque these things off-line to reach high values safely. I used a beefy Proto 30mm crowfoot adapter on a standard torque wrench (compensated for adapter length change), held the shaft in the TTS one way bearing fixture, and then used a second wrench on the tool holder body (end of wrench immobilized by contacting the surface of the workbench) to keep things secure. Worked like a charm, repeatably killing the collet nuts! Lol.



Speaking of High-End ER-32 torque specs... it's between 80 lb.ft. to 125 lb. ft., depending on the manufacturer's rating. With a 4-pin C adapter like this, I can see that happening, but never when clamped in a machine.

-Mike

PS: Picture made me think... looking into it further, that wrench is rated at +/-4% uncertainty... but no mention if that's a full-scale or an "as indicated" value, as is normally given. So, worst case, the wrench could have been (150[lb.ft]x0.04)+57[lb.ft]= 63[lb.ft] from the start, plus error from crowfoot calcs, plus poor technique... reaching. ~67[lb.ft] may not have been out of the question.