What's new
What's new

Dmg mori cmx 1100 v

nsharma

Aluminum
Joined
Nov 1, 2016
Location
Cremona the land of Stradivarius.
Hi,

This is my first email in this forum, have been lurking around for a while tho'.

We purchased a 2001 vintage DMC103v Siemens 810d from an auction in Italy, and then we fixed it up completely. The spindle was replaced, the guides and ball screws were just cleaned and oiled since it didn't have noticeable backlash. Now after using it for a year, it does, and that pains me. The backlash hurts the most in things like drilled-reamed holes. So the options are fix this machine by new/repaired ball screws. Get tool offset and part probing installed, which it does not currently have, and then attach a trunnion of either the CNC type or something like touchdex. I was offered very good money for the machine in December, so something makes me think that this could be an option.

The other option would be to replace the DMC 103v with a new CMX 1100 V with a 5 axis trunnion on the side of the bed. Sometimes we do need the long table and I'm thinking of the option to remove the trunnion for those occasions and then put it back on, for normal work. As everyone in for a new machine, I've browsed almost everything on this forum, including several negatives regarding the DMG-MORI products.

Our main work is related to prototype parts for light aircraft. Most parts would benefit from a 5 axis added trunnion, since they are usually complex and small, some others are very large and would be done without the trunnion on the bed. There are many places where we have holes deeper than 5-7D and I would like through spindle coolant for that. For parts that are made from steel and tough alloys (15CDV6, 4130, 4430, stainless, TiAl), I'll like Minimum quantity coolant/airblast. In Italy, the most used tool holders that are also rather inexpensive are DIN40, of which I already have a fair amount, so the spindle should preferably have that. I'd like to specify glass scales on all axis for long term precision and for in-machine, post work part probing. I'm fairly technical hands on mechanical and aerospace engineer, but I don't have experience with the newest Fanuc/Heidenhain/Siemens controls.

With this in mind, here are some of my initial questions:

  1. Which control would you suggest would be best for my purpose?
  2. Which one of the controls is easier to configure, connect to, extend etc.
  3. Which one of the controls can handle 3D paths better?
  4. I'd like to perform lights out manufacturing, would you suggest a Blum like tool probe or something standard could do the job equally well?
  5. Have you seen the location of part on the Haas UMC750, in which the controller has no issue in setting the part axis away from the trunnion's axis, does any of the above controllers have such a feature?

Thank you in advance.

Naresh

NASHERO
 
Hi,

This is my first email in this forum, have been lurking around for a while tho'.

We purchased a 2001 vintage DMC103v Siemens 810d from an auction in Italy, and then we fixed it up completely. The spindle was replaced, the guides and ball screws were just cleaned and oiled since it didn't have noticeable backlash. Now after using it for a year, it does, and that pains me. The backlash hurts the most in things like drilled-reamed holes. So the options are fix this machine by new/repaired ball screws. Get tool offset and part probing installed, which it does not currently have, and then attach a trunnion of either the CNC type or something like touchdex. I was offered very good money for the machine in December, so something makes me think that this could be an option.

The other option would be to replace the DMC 103v with a new CMX 1100 V with a 5 axis trunnion on the side of the bed. Sometimes we do need the long table and I'm thinking of the option to remove the trunnion for those occasions and then put it back on, for normal work. As everyone in for a new machine, I've browsed almost everything on this forum, including several negatives regarding the DMG-MORI products.

Our main work is related to prototype parts for light aircraft. Most parts would benefit from a 5 axis added trunnion, since they are usually complex and small, some others are very large and would be done without the trunnion on the bed. There are many places where we have holes deeper than 5-7D and I would like through spindle coolant for that. For parts that are made from steel and tough alloys (15CDV6, 4130, 4430, stainless, TiAl), I'll like Minimum quantity coolant/airblast. In Italy, the most used tool holders that are also rather inexpensive are DIN40, of which I already have a fair amount, so the spindle should preferably have that. I'd like to specify glass scales on all axis for long term precision and for in-machine, post work part probing. I'm fairly technical hands on mechanical and aerospace engineer, but I don't have experience with the newest Fanuc/Heidenhain/Siemens controls.

With this in mind, here are some of my initial questions:

  1. Which control would you suggest would be best for my purpose?
  2. Which one of the controls is easier to configure, connect to, extend etc.
  3. Which one of the controls can handle 3D paths better?
  4. I'd like to perform lights out manufacturing, would you suggest a Blum like tool probe or something standard could do the job equally well?
  5. Have you seen the location of part on the Haas UMC750, in which the controller has no issue in setting the part axis away from the trunnion's axis, does any of the above controllers have such a feature?

Thank you in advance.

Naresh

NASHERO

Hi Naresh.

BTW straight off the bat your English is excellent by the way. [Anyway waaaaay better than my Hindi or Italian LOL].

Some indication of ranges of part sizes might be helpful.

Straight off the bat Matsuura MX-520 comes to mind. The reason why I would say that is if I was building prototype light aircraft/flight systems that human beings sit at the controls; IN a cockpit, and it was my designs that were being flown then I'd have a machine that lets me sleep at night. In other words, if an accident or loss of flight system results its not because of what was used to create in house parts assuming good design that mitigates undue stress points and allows for "Creep" in aluminum.

The MX-520 can come with a good sized rectangular table and has a smaller rotary table inset. Good for smaller full five axis parts, (head can get close to the smaller table) but also good for larger 4 axis and three axis parts.

That will set you back around $250 to $300K.

Some indication of tolerances and angular precisions required, size and associated % ages of parts at various sizes and budget might help narrow it down a bit or find a different "Less expensive" option... Or something completely different that might be appropriate.

Cheers,

Eric
 
Hello Eric,

Thank you for your comment and your response.

About aircraft design, if you wish we can sit down and chat next time I visit the US. I've designed a few planes in the past, all of which fly people and one carries paying passengers, used to be flown by AA on commercial routes in the US. I'm also a licensed pilot instructor, I learned to fly in India, soloed in 1982, paying my way by working as a photographer for a film crew in summers. I fly sailplanes, airplanes and have also flown gyrocopters and choppers. The prototype, designed by me, will be test flown by me, very shortly, since almost all of it is already finished. The machine is for the remaining few parts and subsequent production. A bit about my background, in my engineering university, at the age of 21, circa 1983, I designed and built a really tiny IC engine which worked extremely well. I have a photo of it somewhere. It was built on a shaper, lathe, and a very primitive honing setup using a lathe. Very early on in my career as an engineer, I have worked with Studer finish-match grinders, deep hole drilling machines-- drilling 0.2mm holes, 2mm deep in hard steel semi spherical surface, and maho cnc mills and I know how to consistently get a 2 micron taper on a 20mm shaft and how to measure it (using 1986 technology) to know that it is so. So I'm waaaay far from being green around the ears and am quite a nitpicker myself and one will see that in the quality of the plane that is under dev.

Let us keep the discussion to the current machine tools, about which I'm less experienced, other than the DMC 103v + Siemens 810d that I've fixed and use it currently to machine parts. I realized that running HSM paths on this machine caused it to quickly increase its backlash. Or perhaps, I've begun to see its backlash now on parts which was not there earlier. I'm looking for something in the range of +-(8 to 10) microns of precision and a repeatability of around 5 microns in a machine tool after around one-two years of sporadic usage, meaning less than 8 hours/day. About the angular tolerances, I'll like to first get quotes and then see what is the best that a system will provide and then compare it with parts. For most parts with critical dimensions, it is likely that operations on those will be done in the most rigid configuration. I've got no allusions to getting sub 2sec tolerances with these machines. The reason to keep the production inhouse is for the sake of keeping it all under wraps till we patent each patentable aspect. Therefore the machine will not be utilized for all of its capacity, unless someone walks in to my facility with a job request. $250-300k is around 100-120k more than what I'm willing to spend and I think that with what I need, there should be better alternatives, so your rather enticing suggestion of Matsuura MX520, does not bode well with my wallet.

The sizes that we usually deal with differ with the parts. Most parts are small, like 250mm x 100mm x 20, some are a bit larger, but all around that general size. For molds the dimensions could be well over 1000mm, in that case I split up the molds and that works. I've made a lot of tiny molds for fiddly carbon composite parts. I've also machined some toolox on the DMC103v, and it did so with exceptionally high surface quality. Now, that said, the quality of the machine tool is just one aspect, the entire loop is what I need to close, i.e. tool holders (big plus comes to mind), tools, fixturing, table rigidity and the correct feeds/speeds. I've got almost all the rest, dialed in.

I'd narrowed the choice down to the following:

CMX 1100 + Lehmann 2 axis trunnion. DIN40 taper, no big plus, glass scales, blum laser and probe. 5500 kg mass

Quaser mv184P + 2 axis trunnion. DIN40 taper big plus, glass scales, blum laser and probe. 7290 kg mass

Doosan DNM 5700 + 2 axis trunnion. DIN40 taper big plus, glass scales, blum laser and probe. 6500 kg mass.

Huron

All with Heidenhain controller.

So does anyone have suggestions as to the glaring gremlins in the choices above? I was thinking of getting quotes to those and do a proper evaluation. On paper all these look competitive, but I guess field experience is what I need the feedback for.


Thanks in advance.

Naresh

NASHERO
 
Last edited:
Hello Eric,

Thank you for your comment and your response.

About aircraft design, if you wish we can sit down and chat next time I visit the US. I've designed a few planes in the past, all of which fly people and one carries paying passengers, used to be flown by AA on commercial routes in the US. I'm also a licensed pilot instructor, I learned to fly in India, soloed in 1982, paying my way by working as a photographer for a film crew in summers. I fly sailplanes, airplanes and have also flown gyrocopters and choppers. The prototype, designed by me, will be test flown by me, very shortly, since almost all of it is already finished. The machine is for the remaining few parts and subsequent production. A bit about my background, in my engineering university, at the age of 21, circa 1983, I designed and built a really tiny IC engine which worked extremely well. I have a photo of it somewhere. It was built on a shaper, lathe, and a very primitive honing setup using a lathe. Very early on in my career as an engineer, I have worked with Studer finish-match grinders, deep hole drilling machines-- drilling 0.2mm holes, 2mm deep in hard steel semi spherical surface, and maho cnc mills and I know how to consistently get a 2 micron taper on a 20mm shaft and how to measure it (using 1986 technology) to know that it is so. So I'm waaaay far from being green around the ears and am quite a nitpicker myself and one will see that in the quality of the plane that is under dev.

Let us keep the discussion to the current machine tools, about which I'm less experienced, other than the DMC 103v + Siemens 810d that I've fixed and use it currently to machine parts. I realized that running HSM paths on this machine caused it to quickly increase its backlash. Or perhaps, I've begun to see its backlash now on parts which was not there earlier. I'm looking for something in the range of +-(8 to 10) microns of precision and a repeatability of around 5 microns in a machine tool after around one-two years of sporadic usage, meaning less than 8 hours/day. About the angular tolerances, I'll like to first get quotes and then see what is the best that a system will provide and then compare it with parts. For most parts with critical dimensions, it is likely that operations on those will be done in the most rigid configuration. I've got no allusions to getting sub 2sec tolerances with these machines. The reason to keep the production inhouse is for the sake of keeping it all under wraps till we patent each patentable aspect. Therefore the machine will not be utilized for all of its capacity, unless someone walks in to my facility with a job request. $250-300k is around 100-120k more than what I'm willing to spend and I think that with what I need, there should be better alternatives, so your rather enticing suggestion of Matsuura MX520, does not bode well with my wallet.

The sizes that we usually deal with differ with the parts. Most parts are small, like 250mm x 100mm x 20, some are a bit larger, but all around that general size. For molds the dimensions could be well over 1000mm, in that case I split up the molds and that works. I've made a lot of tiny molds for fiddly carbon composite parts. I've also machined some toolox on the DMC103v, and it did so with exceptionally high surface quality. Now, that said, the quality of the machine tool is just one aspect, the entire loop is what I need to close, i.e. tool holders (big plus comes to mind), tools, fixturing, table rigidity and the correct feeds/speeds. I've got almost all the rest, dialed in.

I'd narrowed the choice down to the following:

CMX 1100 + Lehmann 2 axis trunnion. DIN40 taper, no big plus, glass scales, blum laser and probe. 5500 kg mass

Quaser mv184P + 2 axis trunnion. DIN40 taper big plus, glass scales, blum laser and probe. 7290 kg mass

Doosan DNM 5700 + 2 axis trunnion. DIN40 taper big plus, glass scales, blum laser and probe. 6500 kg mass.

Huron

All with Heidenhain controller.

So does anyone have suggestions as to the glaring gremlins in the choices above? I was thinking of getting quotes to those and do a proper evaluation. On paper all these look competitive, but I guess field experience is what I need the feedback for.


Thanks in advance.

Naresh

NASHERO

Yup I have some ideas... Rather busy today and tomorrow and need to do a smidge of "Math" to help with your question.

Here's a simple question for clarification. Are you intending to leave the 2 axis trunions permanently on your 3 axis mill of choice? Is the idea here to achieve cheaper 3+2 applications rather than a dedicated 5 axis machine that is also 3+2 or (4+1). OR are you going to do large "flat work" one week and then another week "Do" 3+2 after dialing in and setting up your trunion?

Straight off the bat (personally) I would not go for a Lehmann two axis trunion if its the one that bolts one (4th axis rotary to another). It's OK for some applications but not appropriate for serious cutting forces and high-ish precision (better than +/- 8 arc seconds repeatability on each axis). DMG Mori make a very good Direct drive DDRT 5ax (type thingy; that's good to about 5 arc seconds on each axis) that is very accurate but weighs 500lbs. Not entirely sure how well that "plays" on the CMX 1100.

Out of the machines you have thrown up there I would be leaning towards a Quaser mv184P There are some very good trunions that are made in Taiwan also.

I still think a dedicated 5 axis machine might serve you better, have you checked out the Quaser MF400c

QUASER MF4C Bearbeitung eines Kettenrades aus 42CrMo4 - YouTube

CNC frasen auf Quaser MV 184P / CNC milling Quaser MV 184P - YouTube (regular mv 184 P).

What Huron(s) new are there that are under $200K?

Also you might want to check out Matsuura MX 330... The working volume maximum part size might fit your main "collection" of smaller parts. Price around $200K and that's sim-5 with a lot thrown in.

I have to admit I would (for me) personally prefer a dedicated 5 axis machine even a Robodrill over a Trunion on a 3 axis machine. The errors really stack up, and I'm not sure I can deal with constantly having to set up and break down a trunion. I'm sure on PM forum here you could get a lot of opinions on how long it takes to set up a trunion on a 3 axis mill.

In our case after researching the snot out of things we realized that we really didn't have any parts that were wider than 10" in Y. So that right now is pushing us very heavily towards B axis mill/lathes and things like a Mazak Integrex I-200 with a 60" between centers. Just kinda fits the nature of our parts tolerances and eventual production rates and workflow. You might stumble across a dimensional "Quirk" with your parts that might push you more firmly in one direction than another.

Another arrangement you might look at is an Okuma M-560V with risers (now offered) and a Koma Precision trunion. You can't get scales on that but the M-560V is really high quality machine where the scales would only improve part to part accuracy but not individual part accuracy. Machine has really good thermal compensation for three axis work, but even with the risers you need to check your Z- clearance. You could probably get something put together for you around the $200K mark.


AND finally check out the Hardinge Bridgeport XT630 5ax. That machine will do all your parts for about $260K. Value for money that's hard to beat and you can get it with a Heidenhein control.


Generally I try to look long term and not have to successively junk three generations of machines to where I really need to be but rather (if possible) cut out ten years of BS and waste and try to go for the machine that is really needed in one go. Daunting as that may seem long term that's a more cost effective and waaaaay more efficient. Financing of one sort or another can help you with that.
 
Hi Eric,

I don't have my computer with me at this moment. I felt compelled to answer your questions. I do need the 1000 x 600 table often and the +2 axis will be mounted when I'll be doing the many tiny parts, and will be running unattended. I'm not a fan of heavy cuts, and prefer high speed, constant chip load, full flute engagement cuts. My pet peeve with my current setup is the inability to do lights out machining with confidence, and the necessity to baby the machine. I blame the geriatric controller for that.

Ideally... dreaming loudly, I should have a very large 5 axis gantry mill and a high speed multi spindle, multi head, multi axis, turn mill... but I can only afford these punny machines ;)

Btw the lehmann table I'm thinking of is ddm and is fully supported 4th and the 5th is a bit unsymmetric towards the 4th's drive side. It is fast and claimed to be much better than 5 arc secs.

Sent from my GT-N5100 using Tapatalk
 
Hi Eric,

I don't have my computer with me at this moment. I felt compelled to answer your questions. I do need the 1000 x 600 table often and the +2 axis will be mounted when I'll be doing the many tiny parts, and will be running unattended. I'm not a fan of heavy cuts, and prefer high speed, constant chip load, full flute engagement cuts. My pet peeve with my current setup is the inability to do lights out machining with confidence, and the necessity to baby the machine. I blame the geriatric controller for that.

Ideally... dreaming loudly, I should have a very large 5 axis gantry mill and a high speed multi spindle, multi head, multi axis, turn mill... but I can only afford these punny machines ;)

Btw the lehmann table I'm thinking of is ddm and is fully supported 4th and the 5th is a bit unsymmetric towards the 4th's drive side. It is fast and claimed to be much better than 5 arc secs.

Sent from my GT-N5100 using Tapatalk

Do you have a link for the specific Lehmann table you are talking about.

You mean like this?

Lehmann 5 axis CNC rotary table on an XYZ machining centre - YouTube

I thought the Lehmann tables were/(used to be) of the order of 20 arc seconds as a 4th axis. +/- 15 arc seconds at least for repeatability (off the top of my head) so I certainly didn't mentally (in the past) put them in the category of better than 5 arc seconds) but maybe the new anti backlash gears they are using are closing significant "Slop". Still wonder about how orthogonal they are given their arrangement. If they really claim less than five arc seconds on those you should ask to see actual test sheets and plots on those so you get a better statistical feel for various confidence limits. There are very few Trunion based systems and machines that really operate truly below 5 arc seconds... (especially without scales on rotary axes) But maybe Lehmann are really doing something here (direct drive as you say)? :-) On other machines five to 2 arc seconds is the domain of Yasda and Okuma MU-V series type machines as well as dedicated Makino 5 axis machines (that don't plonk a trunion) on a F3.

If that is true and it does what it really says it does then maybe a CMX-1100V wouldn't be so bad given the other linear tolerances you have stated.

Do you have a test part/ (series of parts) that would unequivocally answer whether that is a good fit for your application?

I'd be interested, and I assume you have good QC methods to test that?
 
Hello Eric,

This is immensely useful. You have hit the nail on the head, yes its that type of a table. MTD have a few videos on Lehmann. They come in three flavours and all of them have encoders on the axis. The accuracy is claimed to be +-2 arc sec. Well, I'll like to know if someone has them and if this is indeed true. The ddm type is Vario and Vario X1. There is a great review of all the +2 axis here: Quality turn | Practical Machinist

The tables that interest me are the bottom two in the list: T1511520 and T1520520.

image3034.jpg

The type of parts that will be machined are very varied, mainly things like bellcranks, hardpoints, inserts, rod-ends, couplings, control system parts, hinges, door-levers, AN fittings, small valves etc. I can send an image of something that has been made on the DMG 103v of mine to give an idea. But all the small parts will benefit from 5 axis machining. They are fairly complex and many times need multiple setups for the same side due to various reasons. I usually do all the accuracy related stuff such as drilling reaming, or accurate lands for bearings etc., if I can, on prismatic shapes, then go on to machine the contours, finally do the lightening cavities etc. All taps are roll taps and they work well.

So here is a typical part and complexity of what I do.
2016-04-18 14.26.15.jpg

This began as a 6000 gm of Al alloy, and ended up 268gm. Its size is very typical of most parts.

About Huron, the VX10 is around the same price range as the one's I've shortlisted. They are around 6700 kgs.

I just had a visit from the Doosan guys, and will go to see their machines next week, he told me to consider a DNM6700 rather than the one I was looking at i.e. DNM5700. He didn't seem to be a fan of Heidenhain and also not of the linear scales. He said that to keep them clean is very difficult, as soon as the rubber wipers are gone, there is no way to pressurise the inside of the scale with air, due to now leaky seals. This then becomes the beginning of the end of those scales due to contamination. They have very fancy thermal control of the structure and the spindle, but not of the ball screws or guides. They have 15000 machines installed in Italy alone!!! He said that, in his client base, he's never seen complex stuff as I do, but thought that his machines are much more suited to the job and are less expensive than the rest... ;-) obviously, I thought!

I'll like to hear from people regarding their experience with these machines. While an all kitted out DMG looks more like a kitchen refrigerator, the Quaser and Huron look like mean workhorses, the Doosan a regular machine... I'm still not sure which will last longer, cost less to run, hold tolerances, handle high speed trochoidal machining without hiccups, easier to maintain, have fewer gremlins and generally have a better resale price.

Thanks a lot,
Naresh

NASHERO
 
Last edited:
Hi Eric,

I just checked the play in the dmc 103v, its 26microns (0.026mm) in X and 62 (0.062) microns in Y. Z, being a heavily preloaded axis does not have any noticeable play.

I'm going to ask a regrinding service to regrind the two X and Y screws and then replace the bearings, at least till the new machine is ordered and installed, I'll be ok with work.

Best,
Naresh

NASHERO
 
Hello Eric,

This is immensely useful. You have hit the nail on the head, yes its that type of a table. MTD have a few videos on Lehmann. They come in three flavours and all of them have encoders on the axis. The accuracy is claimed to be +-2 arc sec. Well, I'll like to know if someone has them and if this is indeed true. The ddm type is Vario and Vario X1. There is a great review of all the +2 axis here: Quality turn | Practical Machinist

The tables that interest me are the bottom two in the list: T1511520 and T1520520.

View attachment 190701





NASHERO


Hi Naresh,

I'm going to have to cut this one into separate parts. I'll get to the other parts later/tomorrow.

Looking at the graphic/except from the Lehmann catalogue... That may be the unsupported one (at one end mechanically).

I looked up their current online catalogue and some of the values are a little different.


Just as mild "Sanity check" on the table you cite/post. [I'm not bashing Lehmann and they are very nice for a range of applications].

I invite you/entreat you to put on your design hat. Clearly with the design work you have done with "Augusta" (beautiful elliptical wings BTW) if you yourself were tasked to design and build a 5 axis unit to work on a three axis mill, you would never in a million years design it the way Lehmann has.

Basically the Lehmann rotary looks beautiful (nicely made and decent price point) but really is intended for medium to lowish precision work (originally).

So if you look at the table of numbers you posted it states: Standard Indexing accuracy for the Trunion (on the T1-520520 ) of +/- 45 arc seconds. That's a spread (probably to three sigma of 90 arc seconds or 1.5 arc minutes. With a gear improvement that increases to +/- 41 arc seconds. So you will definitely (as they advise) have to employ pitch error compensation on the control to compensate elastic deformations and eccentric loading etc. The rotary table has stated +/- 12 arc seconds... So that's a spread of 24 arc seconds and better back lash and tighter gears is reduced to +/- 8 arc seconds, a spread of 16 arc seconds. The newer model on their catalog for example T1-511520 states for indexing accuracy (standard) as +/- 16 arc seconds on trunion axis (32 arc second spread) and +/- 17 arc seconds for the rotary. [It's a little odd how they break out 4th vs 5th].

So you have to turn to page 35 for an explanation of how they derived those values and then things get a little weirder. LOL.

So you can indeed bolt on as an option a Heidenhain encoder. The ones they use are very expensive (at least to buy off the shelf) and are very bulky compared to Renishaw equivalents. So that can give you your +/- 2 arc seconds BUT more in terms of "Sensitivity" which is quite different from actual precisions and actual indexing accuracies. So the encoder can "Report" to your machines control, BUT the control has to interpret that value and the actual move has to munge through pitch error compensation as well as mechanical errors to perform the movement not least of which is/are breaking errors. I'm not sure that there is a lot of "Servoing" for finding the near correct indexed position and then "break accurately" or you keep the whole system holding in dynamic mode. To my eye practically I'd be surprised if those systems actually position better than +/- 10 arc seconds on the rotary and +/- 15 arc seconds on the trunion WITH the nice Heidenain encoder (and that's to a two sigma confidence limit) :-)

I would call Lehmaan out on that as they design beautiful products but they seem to have "Cheaped out" on producing a NEW dedicated five axis trunion unit, and instead have cobbled together almost unaltered two sizes of 4th axis units which in a sense violates their beautiful design of their 4th axis.

Their brochures are beautiful but the full explanation of volumetric accuracy and the tables they provide, they say were carried out on a (DMG) DMU 65! (I'm like (WTF man!)). So the table seems to be of a completely different machine (a universal). That's on page 81. So you have to rifle through three sets of references on separate pages and you finally arrive at a dead end and irrelevant /non-explanation (in other words they totally side step what their product really does BUT they do provide some nice mathematical/ theoretical tables). When companies like Yasda in their brochures quote positional accuracies and precisions and repeatabilities for rotary axes they provide multiple pass plots of what's going on, so that you have absolutely no doubt as to what the machine is capable of.


@Naresh "Dude"/ Dr Sharma you and your planes and your company deserve better than that.

For 4th axis Yup its pretty good / nice, but for real 5 axis 3+2 a Koma precision or DDRT 5ax (fro DMG mori etc) or more coherent designed trunions would probably serve you better.

I think when you add up the process improvements of the Lehmann + better gears + lots of tweaking and compensation at the control + expensive and bulky Heidenhain encoders its probabaly going to get close to $50K any way and for $42-65K can get you a really solid and reliable 5 axis trunion.

That's kind of why I'm more of a fan of dedicated 5 axis machines as they are tweaked and tuned to do exactly what they are supposed to do... Even a Robodrill five axis arrangement or a Mill Tap 5 axis from DMG mori.

You might want to have a look at a 5 axis DMG Mori "Mill-Tap" and their robot arm / work piece swapper/ cell arrangement , (as you mention the lights out "idea").

Just as random thing how is MAZAK in your area? (Is it well supported and represented? Who in your areas is the strongest for local presence and support in your opinion)?

So that's piece one ... LOL

[I used to work for the National Air Space Museum almost 20 years ago as a conservator and materials scientist... (We had the Enola Gay in 50,000 pieces at that time, and I also worked on space suit/ modern materials research/ Deterioration... But later switched to computer science and 3d imaging... Kinda interesting that we both have "cross over backgrounds" and a passion for flight but also insist on being very hands on. I used to be part of the RAF Cadet corps (good way to get in a lot of free flying LOL). To be honest I most like to fly the cloth and stick aircraft, but your airframe on Augusta looks amazing (as much as we are able to see) and the safety features seem excellent; so hoping one day to see these flying around Denver.)
 
Hi Eric,

I just checked the play in the dmc 103v, its 26microns (0.026mm) in X and 62 (0.062) microns in Y. Z, being a heavily preloaded axis does not have any noticeable play.

I'm going to ask a regrinding service to regrind the two X and Y screws and then replace the bearings, at least till the new machine is ordered and installed, I'll be ok with work.

Best,
Naresh

NASHERO

How do you know it's the screws & ball-nuts? Typically the thrust-bearings will wear/fail before the screws & ball-nuts will. Also, if the slop is in the thrust-bearings, then your backlash measurements should repeat, at any point in that axis' travel. If the wear is in the screws, then I would expect there to be more backlash in the center of travel, but much, much less at the far extremes near the over-travel limits.

Just a thought before you spend money on a potentially unnecessary repair.
 
@Jashley73 thank you for reminding me. I will check the entire surface in 100mm increments and see if it is indeed from the thrust bearings or the screw. Actually, when I purchased the machine it had been sitting idle for over 3 years, the old owners, who's company had gone bust was very nice to me and warned me of the spindle and said one axis may require regrinding. I changed the spindle, but didn't do the axis since I was in a hurry to get stuff finished. Now that you mention, in retrospect, I should've done that.

Actually I remember that close to the edges, there is less backlash. I had to use the whole table thus noticed the effect on parts in the center. But I have no way to understand if the thrust bearings are also a contributor other than by opening the way covers. Do you have any suggestions?

Today I had some more visits, one company suggested that they could refurbish the machine and upgrade the controller to 840d that will eliminate the issue with trochoidal machining. I have no clue how much will that cost, since at the moment, I don't have time to spend on the machine, it has to either be completely fixed or replaced.

best,
Naresh

NASHERO
 
Last edited:
Hello Eric,

Thanks for dissecting the Lehmann stuff for me. It is indeed illuminating that the FUD is hidden from the user under the crafty marketing presentation. As I said, I'm not in the market for a yasda or a mitsui. Looking at the end supports that leave a lot of questions unanswered, regarding the X-Y stiffness of the tall rectangular supports, I can imagine that deflection cannot be avoided. I just thought that the accuracy and repeatability results in the two support (check out page 27 of the catalog) were measured using some DIN standards and would be reliable. As for now, with our parts, I'll not even approach the loads that they claim. The lehmann interest arose from the fact that DMG will have one sitting on a machine at the openhouse.

Yesterday, the doosan guys suggested to use a koma precision table as well, it would end up being a 35k upgrade. Next week, I've got folks from Tsudakoma coming into the facility with their offer. Well, I'm hoping to learn a bit more about these solutions from them. The Tsudakoma said that their tables are made from a single grey-cast-iron body, so the support for the 4th is solid, and the whole table can be moved easily as a single piece. We'll see how it ends up.

I looked up some 5 axis machines during this search, one was a yasda that was available 2nd hand, almost 15 years old, but from a third party precision report, it still held between 1.3-2microns of precision, on all axis, with a refurbished spindle. The other was a hermle from 2013. The yasda service folks in Italy are a bunch that I'd be happier to avoid. The hermle had done 42000hrs of on-time, and 26000 hours of spindle time in 6 years! Thats equivalent to one working lights out 24/7 for its entire 6 years of life. So I think that this hermle has had a wild ride and, without tools to ascertain how much it has been consumed or damaged with such usage, I don't feel comfortable to pledge around 200k. Then, again, if its still good, i beg to question: why do they want to retire it early?

WOW, Smithsonian! I've visited that often. Didn't spend much time with Enola Gay, what a beautiful aircraft, but what a tragic part of human history. My dad was a scientific glass blower, he made gas lasers, one of the first ones in the world in 1960's, at the MIT's Lincoln labs. We had original photos of the lunar landing at our home in India, they were given to all those who worked on the lunar landing mission, which was orchestrated by the Lincoln labs. So visiting Smithsonian to see the apollo program's relics has always held a special place in my heart. Thank you for your highly appreciated comments on Augusta. She's the fruit of love, passion and extremely hard work. RAF Cadet's, really! That's cool. We should definitely have a few beers on my next visit and keep personal conv. out of the forum as, I'm sure its a boring diversion for others.

Naresh

NASHERO
 
Doosan DNM 6700

@Naresh

I think you will like the DNM 6700...

Doesn't that sound patronizing??? A lot of sales people try to TELL you what YOU will like LOL.

But seriously the Doosan machines are pretty terrific, just stick your fingers in your ears when they try to tell you that the Doosan models are just like Mori Seiki of yesteryear. IMO They/Doosan are pretty terrific/good machines in their own right.[Sometimes like all machines there's some random teething troubles like a tool being thrown and smashing a window, but that's rare, little quirky details that can afflict any machine].

Sounds like with the Doosan guys in Italia that you should be able to close in on a workable solution for you that would serve you well. Those machines do last a long time are pretty solid and retain decent value and resale value and hold good tolerances over a long period. [If I was going to go up from there in terms of higher precision spindle then I would check out Matsuura VX 1000 or VX 1500]. Good to have another machine to compare against that ultimately might make you feel very good/better about a possible Doosan purchase. ON the VX's you should be able to get scales and the spindle IS/ARE at the edge of being sub micron effective run-out... But may not be quite as robust as Doosan for hogging out a lot of material but comparable.] If you really wanted to save $$$$$ on a "Hunky" yet pretty precise machine check out the Hardinge V1000; technically the Mits control should be capable of 3+2 and 4+1. That might give you another machine to compare with... [Quite similar to the Quaser that you were looking at originally]. Nice to have choices and things to choose between, right? Or drive one mad LOL :-)].

The Doosan models offered in the USA are different from EU (what's left of it).

Conceivably DNM 6700 would have a table size (1500 x 670 [59" x 26.4"]) <Travels 1300/670/625 mm> that might allow you to have a small 5 axis trunion hanging off one end of the table almost permanently... I don't know your exact size of parts, but that could be quite sweet with a couple of other vices you can get to your "Sixth" face/ top and tail from 5 sided, as well as still have enough room for longer parts.

I'm assuming you are doing some mold work for your composite airframes etc. and when you use the phrase "Lights out" you mean you can leave the machine unattended for extended periods without "jacking-up" a mold that might take days to complete?

Also sounds like from the way you like to "cut" that high spindle speed is going to be important to you. So the 15K spindles might be more attractive than 8 or 12k. I think on some models (with Doosan) that switches over to the 650HS that might have a different table size and travels.

In this case I'm not sure about the need for a Blum laser tool setter. Those add about $25K ; they are good but usually seen on machines with 60 tools and cell arrangement/high degree of automation.

As far as your comment about the control. Typically Doosan handle more Fanuc based machines in the USA; so typically that means machine tool sellers/vendors are a little weirded out by the Heidenhain control. Doesn't mean the Heidenhain control is bad; typically a very good control for higher end five axis applications. Interesting that Dooosan also offer Siemens control; given that you already have experience of Siemens you might like that. Also make sure the control can actually handle the size of programs you want to run now and in the future.

Given that there are 15000 Doosan machines in Italy, see if you can find a shop that is doing similar 3+2 work to the standard you need. Ask the Doosan sales guys how long it takes to set up a two axis trunion and dial it in and (hopefully) have some no brainer kinematic calibration cycles. Ask them about dynamic work offsets as well.

Scales? I have heard similar things. That's one of the nice things about the Magna scales on the Mori Seiki machines/ DMG Mori is that they are not so fragile and susceptible to contamination (allegedly).

You are not going to have any trouble with trochoidal/HSM machining strategies on a Doosan 6700, or Matsuura VX or Hardinge V1000 or Quaser (equivalent) machine.

So if you don't mind me asking are you using CAD/CAM or are you using your own software to generate some of the G-code for the parts you are making?

Also do you have some sort of climate control in your shop or are you working out of an aircraft hanger?


[I'd still try to twist your arm one day for a sim five axis dedicated machine. For me I like the possibility of deburring/chamfering along complex edges even if I am doing 3+2 principally (saves a lot of hand work and QC problems as well metal finishing). Also a really decent five axis machine (doesn't have to be Yasda or Matsuura) can accomplish some or most of your higher tolerance surfaces and procedures that right now you are doing by hand. Time IS time... ]
 
@Naresh

Mills CNC show off titanium 'cut of death' at Nikkens Beyond Innovation - YouTube

Purely by accident stumbled across this... Shows a DNM 5700 with a Nikken 5 axis trunion table on the end of the table doing pretty similar work (by the sound of it) from what you have described.

Personally I think the Koma Precision tables are smidge higher tolerance (newer more interesting options these days) (same with Tsudakoma... Potato Potaaaato) but the Nikken rotaries are very robust and take a serious pounding day in and day out for years.

Cheers
 
Hello Eric,

Thanks again for your help.

@Naresh

I think you will like the DNM 6700...

Doesn't that sound patronizing??? A lot of sales people try to TELL you what YOU will like LOL.

But seriously the Doosan machines are pretty terrific, just stick your fingers in your ears when they try to tell you that the Doosan models are just like Mori Seiki of yesteryear. IMO They/Doosan are pretty terrific/good machines in their own right.[Sometimes like all machines there's some random teething troubles like a tool being thrown and smashing a window, but that's rare, little quirky details that can afflict any machine].

Sounds like with the Doosan guys in Italia that you should be able to close in on a workable solution for you that would serve you well. Those machines do last a long time are pretty solid and retain decent value and resale value and hold good tolerances over a long period. [If I was going to go up from there in terms of higher precision spindle then I would check out Matsuura VX 1000 or VX 1500]. Good to have another machine to compare against that ultimately might make you feel very good/better about a possible Doosan purchase. ON the VX's you should be able to get scales and the spindle IS/ARE at the edge of being sub micron effective run-out... But may not be quite as robust as Doosan for hogging out a lot of material but comparable.] If you really wanted to save $$$$$ on a "Hunky" yet pretty precise machine check out the Hardinge V1000; technically the Mits control should be capable of 3+2 and 4+1. That might give you another machine to compare with... [Quite similar to the Quaser that you were looking at originally]. Nice to have choices and things to choose between, right? Or drive one mad LOL :-)].

The Doosan models offered in the USA are different from EU (what's left of it).

Conceivably DNM 6700 would have a table size (1500 x 670 [59" x 26.4"]) <Travels 1300/670/625 mm> that might allow you to have a small 5 axis trunion hanging off one end of the table almost permanently... I don't know your exact size of parts, but that could be quite sweet with a couple of other vices you can get to your "Sixth" face/ top and tail from 5 sided, as well as still have enough room for longer parts.

Yes it sounds rather patronizing and makes one think what on earth is going in the haystack head of the sales guy.

The DNM5700 is the right size and although I do have parts such as molds that can sometimes grow upto 1m in length, as I said I can divide them in parts. I'm still not sure that seals etc. will blow away, since the heidenhain guys and the Quaser guys told me that their linear scales don't sport rubber wipers.

I'm assuming you are doing some mold work for your composite airframes etc. and when you use the phrase "Lights out" you mean you can leave the machine unattended for extended periods without "jacking-up" a mold that might take days to complete?

You are absolutely correct here as well.

Also sounds like from the way you like to "cut" that high spindle speed is going to be important to you. So the 15K spindles might be more attractive than 8 or 12k. I think on some models (with Doosan) that switches over to the 650HS that might have a different table size and travels.

As far as spindle is concerned, I'm looking for a 15k or 18k inline cartridge spindle as opposed to an electro-spindle.

In this case I'm not sure about the need for a Blum laser tool setter. Those add about $25K ; they are good but usually seen on machines with 60 tools and cell arrangement/high degree of automation.

That's an interesting point. I'm planning on setting up a robot to tend the machine, tools and some other silly tasks. It will either be a modified commercial robot, or something designed by me. So either a laser or a touch based sensor will be very useful, but its not fundamental.

As far as your comment about the control. Typically Doosan handle more Fanuc based machines in the USA; so typically that means machine tool sellers/vendors are a little weirded out by the Heidenhain control. Doesn't mean the Heidenhain control is bad; typically a very good control for higher end five axis applications. Interesting that Dooosan also offer Siemens control; given that you already have experience of Siemens you might like that. Also make sure the control can actually handle the size of programs you want to run now and in the future.

The size and complexity of the programs that CAM spits out is humongous and that is where my current 810d begins to show its weakness. Even on simple curves, I can identify facets.

Given that there are 15000 Doosan machines in Italy, see if you can find a shop that is doing similar 3+2 work to the standard you need. Ask the Doosan sales guys how long it takes to set up a two axis trunion and dial it in and (hopefully) have some no brainer kinematic calibration cycles. Ask them about dynamic work offsets as well.

Next Tuesday, I'll be visiting them and will be able to get some answers. They have also promised to show some other shops. But the same has been told to me by the Huron, Quaser and DMG folks, and I'm not worried of info-overload, but I'll be more than happy to learn what they have to say.

Scales? I have heard similar things. That's one of the nice things about the Magna scales on the Mori Seiki machines/ DMG Mori is that they are not so fragile and susceptible to contamination (allegedly).

You are not going to have any trouble with trochoidal/HSM machining strategies on a Doosan 6700, or Matsuura VX or Hardinge V1000 or Quaser (equivalent) machine.
So today the Quaser guys were here. The quaser guys wanted to listen to the spindle of my DMG-103v and were impressed that it was quite with no vibration at its max speed. They will send me an offer for the 184p and, based on how pleased they seemed to be with the DMG 103v, they'll give me something close to what it cost me. While I studied their reactions, it made me reflect-- why not fix this machine a bit more and just use this (103v), its built like a tank. Regrinding the ball screws and new balls and thrust bearings will cost around Eur550/axis plus two days of installation work (for all axis). So the worst case is Eur2400 for all the three axis.

So if you don't mind me asking are you using CAD/CAM or are you using your own software to generate some of the G-code for the parts you are making?

I used to use Rhino3D and spit out my own codes, most of the aircraft has been machined with that method. But with the metal parts, I'm using Fusion3D ultimate. That is 4 axis continuous and with 5th axis swarf milling. The post processor for 810D has been written by me based on the generic Siemens post on Fusion3D. My 3D solid modeler is Solid Edge, but the surfaces are made from Rhino3D. If I see that as a limitation, we'll get NX-CAM.

Also do you have some sort of climate control in your shop or are you working out of an aircraft hanger?
We're in an industrial facility, not a hangar. We have some areas that are controlled, including a clean room for laminations, but the 103v is not in one of them. We'll be moving out of there by the eoy 2017, then we will have a temp-humidity controlled shop space for CNCs and a cutting room, clean room, prep room etc. We're applying for AE9100 and nadcap certs so everything in the facility will be squeaky clean.

[I'd still try to twist your arm one day for a sim five axis dedicated machine. For me I like the possibility of deburring/chamfering along complex edges even if I am doing 3+2 principally (saves a lot of hand work and QC problems as well metal finishing). Also a really decent five axis machine (doesn't have to be Yasda or Matsuura) can accomplish some or most of your higher tolerance surfaces and procedures that right now you are doing by hand. Time IS time... ]
If there was no financial constraint, I'd have done that in a snap. What are your views on pre-owned machines? Did you read the offer that I recently had for the hermle?
 
Hi Eric,

This is exactly something I've been thinking of as my current compromise. I had already seen it a few days back, this is what made me think of calling the doosan folks too.

Still in the evaluation process, I'll look at all the options before stepping over.

Wish you a wonderful weekend.

Naresh

NASHERO
 
nsharma said:
The size and complexity of the programs that CAM spits out is humongous and that is where my current 810d begins to show its weakness. Even on simple curves, I can identify facets.

That may just be the back-lash in your screws/thrust bearings. If it's faceting at the quadrants, then it might just be the backlash showing up in the quadrants, as the axis stops & reverses direction. There should be parameters for backlash compensation, which would help with faceting at the quadrants.

I believe you asked earlier on how to isolate the slop to the thrust-bearings and/or screws... Pull back the way-covers, and find the end of the ball-screw. Put an indicator on the end of the ball-screw, and push against the table in that axis. Push in both directions, and watch your indicator. This should help isolate any backlash that's in the thrust bearings, vs. what you see when you put an indicator on the table/part, where you would see the combined back-lash of the thrust bearings + ballscrews.

If you tighten up the backlash to a few microns, you'll see the faceting on arc-moves greatly diminish. Get the backlash compensation parameters dialed in, and the faceting may disappear altogether.
 
Hello Eric,

schhhhhnip <>


I looked up some 5 axis machines during this search, one was a yasda that was available 2nd hand, almost 15 years old, but from a third party precision report, it still held between 1.3-2microns of precision, on all axis, with a refurbished spindle. The other was a hermle from 2013. The yasda service folks in Italy are a bunch that I'd be happier to avoid. The hermle had done 42000hrs of on-time, and 26000 hours of spindle time in 6 years! Thats equivalent to one working lights out 24/7 for its entire 6 years of life. So I think that this hermle has had a wild ride and, without tools to ascertain how much it has been consumed or damaged with such usage, I don't feel comfortable to pledge around 200k. Then, again, if its still good, i beg to question: why do they want to retire it early?



What are your views on pre-owned machines? Did you read the offer that I recently had for the hermle?

Kinda interested in what model and machine of Yasda? And how much are they asking?

[Are the service guys for Yasda out there really that bad or just kind of aggressive and abrasive? Probably Yasda would be in government/Euorpean/Nuclear/non-civilian/aviation type environments... I've worked in Italy and with Italian crews and always had a real blast but I have also read really funny and bad things (written by Umberto Eco about the Italian Civil service); maybe it's machine tools meet Italian Government work LOL].

Given that a larger percentage of Yasda machines are typically "Mother machine" type accuracy and have a lot of subtle knowledge of the "Art" to make something that's still pretty damn special, they are always worth a look. Check out threads by Dstryr (Dennis) a really smart and forward thinking young machine shop owner just purchased a $600K second hand 5 axis Yasda platform (that's his third or fourth 5 axis purchase I belive)... He has a really sweet (spindle farm).Yasda may be complete overkill for the tolerances you really need but on the other hand if one was close to me for a good price that would be no-brainer.

As far as the Hermle goes... which model and more critically what were they cutting most of time? Sometimes you do get lucky and you find machines that have pretty much been cutting "Plastic" all their life.

The general topic of second hand machines is VAST and impossible to do justice to as a topic in one post. There are literally thousands of threads on PM forum discussing the specific ins and out of particular machines as well as more general cases.

My opinion? (as you ask) :-0. There are no magic beans for second hand machines.

You get exactly what you pay for. If you want a very in depth analysis of what "Older" machines cost you in the long run over newer faster more capable machines comb through the vast library of Webinars from Makino; You'll find those on their web site, but not on youtube (and you have to register and login but is really worth it if you want to learn more).


My take on second hand machines is that it entirely depends on your business, your business model and the trajectory of your business over then next three, five, ten to fifteen years. Because at some point you are going to have to replace your machines (less of an issue perhaps with Matsuura and Yasda and the likes perhaps).

What I have done is to find the main and principal brokers of second hand machine tools, and just spend some time to get to know them and get pre approval for financing squared away in advance (if relevant). Sometimes you do see machines that are gems or perfect fit for your applications and business and you just have to have the ability to move very quickly (i.e. have a "war chest" at you disposal that you easily pull the trigger on things (subject to inspection/ball bar tests and some sheet metal removal/inspection etc. etc. ). Typically (like buying horses) you win some you loose some but over a five year period if you are wise your good purchases out weigh your lemons (that's appropriate for a big outfit with a lot of machines). BUT... If you are only buying one or two machines as the cornerstone of a critical aspect of your business then I would buy new (in this type of case). The only regrets I have of the past 12 years or so is actually underinvestment in equipment vs over investment. If you underinvest and are stuck with equipment, that is not so great, that can cause more problems in the long run than having something that is more capable NOW, but costs $150-200K more.

Business model and intended business direction is (I think) and HOW you really want to spend your time over the next ten years, is the most important factor for deciding what type of investment and machines you need. For example if you only need to buy machines only to have proof of concept and develop and prototype a small number of aircraft, where the business model is to attract further investment of the order of tens of millions then buying second hand machines for POC IS (theoretically and practically) viable; simply because the goal is to attract the next level of investment so that you can buy whatever machines you want in the near future. IF on the other hand the machines that you need to purchase are one of the main planks to transition from prototypes to genuine in house manufacturing and you need to develop serious competency with five axis systems then I would be inclined to buy far more serious equipment up front... That's what we/my company is doing as we have to roll out very high quality items that are not typical for "Civilian" applications" (High mix low volume, high quality, high complexity parts for our systems).

I think there's another hidden factor related to time that folks don't really talk about very much and that's the time taken to master your craft (whatever it may be). The ten thousand hours principal really applies here. Yup in your first year you can certainly crank out a few good parts but overall with five axis stuff it's really going to take about ten years to become "fluid and one" with a range of 5 axis processes in a highly efficient way so that it becomes almost instinctive and second nature. I think there's a hell of a lot of different and important techniques that exist between so called 3+2 to sim-5 axis. So in a sense I have built in budget to learn snot out of all of these things and the value of that learning process... That does have a price and you can assign a $ value to that. Really depends on where you have clear plan and vision for exactly where you want to be in ten years time and also have a lot of viable plan Bs so that you can pivot and be flexible enough to turn your hand to slightly different applications and industries if market conditions change dramatically (either for positive or negative reasons). So that "insurance" factor (if you like) is another reason to invest in really good equipment. If you have investors then they don't always understand why you spent $300K on second hand machine that looks like hell (and makes weird clunky noises) as opposed to a new bright shiny new machine that's actually not as good as the "Ugly" machine. Personally I think the machines that you purchase to manufacture your products with should reflect the quality, precision and sophistication of the items you make.


Hope that helps; but PM forum is vast lake of amazing information on all of these types of matters.

Cheers,
 
cameraman said:
Personally I think the machines that you purchase to manufacture your products with should reflect the quality, precision and sophistication of the items you make.

I love this... I wish more people had this same philosophy. I realize that they're just tools in many cases, but nothing inspires me to do good work, like getting to do the job with good tools.
 








 
Back
Top