What's new
What's new

Advice needed, new high precision machine

ricardo_gt

Aluminum
Joined
Aug 5, 2012
Location
Portugal
Hi folks,
My company signed a new deal to fabricate some hard toleranced parts at 2020 .
Parts need to be guaranteed with several h6 tolerances and some with N1 rugosity.
We came to conclusion that we can achieve this results better if we invest at a new pair of cnc machines.
We are looking to yasda or kern to fabricate the CNC machines that we will need. However, as allways, some advices are welcome.
We have controled temperatures at hour shop and severall CNC machines working fine: 5 axis hermle Mills and Hyundai wia CNC lathes.
We program all the machines with mastercam, but, lathes and Mills have different controles, fanuc on the lathes and heidenhain on the Mills. I am a litle unsure if the new machines should been controled to one or other control, 5 axis. We have around 1.000.000$ budget excluding tooling.
What kind of advice can you give? Thank you!
Best regards!
 
I can't exactly say what parts are, sorry.
Stainless steel high temperature resistant.
Tubular and flat shapes with several grooves. Outside and inside threads. Parts must perfectly match each others.
Tubulares around 120mm long and squares around 150*200*45.
Thank you.
 
Hi folks,
My company signed a new deal to fabricate some hard toleranced parts at 2020 .
Parts need to be guaranteed with several h6 tolerances and some with N1 rugosity.
We came to conclusion that we can achieve this results better if we invest at a new pair of cnc machines.
We are looking to yasda or kern to fabricate the CNC machines that we will need. However, as allways, some advices are welcome.
We have controled temperatures at hour shop and severall CNC machines working fine: 5 axis hermle Mills and Hyundai wia CNC lathes.
We program all the machines with mastercam, but, lathes and Mills have different controles, fanuc on the lathes and heidenhain on the Mills. I am a litle unsure if the new machines should been controled to one or other control, 5 axis. We have around 1.000.000$ budget excluding tooling.
What kind of advice can you give? Thank you!
Best regards!


Truly fascinating question.

What is a N1 "Rugosity"

Rugosity - Wikipedia ?

Is that like a topological version of Ra or Rz (does that surface energy and texture apply more to high surface area parts like 3d printed metalics ?).
 
I can't exactly say what parts are, sorry.
Stainless steel high temperature resistant.
Tubular and flat shapes with several grooves. Outside and inside threads. Parts must perfectly match each others.
Tubulares around 120mm long and squares around 150*200*45.
Thank you.

We / I have very similar application requirements (superficially).

Without knowing more about your form tolerances and how one set of surfaces may have to relate to another set orthogonally or on weird angles it's hard to give 'Accurate" advice. So in sense more randomly applied advice is given (no change there lol).

First thoughts that comes to mind are...


1. Cycle time and automation-ish / multitasking ?

i.e. ---> How many of these in a year do you have to make and what kinds of efficiencies are required to make what you are producing viable from a business point of view ?

2. Are vertical mills really the answer ? [ what a bout B axis mill turn machine ? (sim 5 axis capable ?)].

Something like a 5 axis (B axis) mill turn machine can be very accurate . Typically for turned surfaces a turning operation on a good turning center can be sub-micron roundness and form tolerances of the order of 5 micron or less.

Also where you have circular or tapered features that have to be orthogonal to referenced accurate longitudinal milled features;, depending on skill , fixturing and general process awareness high part accuracies can be achieved in an efficient manner. Same / similar for large bore threaded features and precision threads that may be part of a more complex mill-turn type geometry. Those being good reasons for a 5 axis B axis mill turn machine.

IMO form tolerance of turned features is much higher than that of vertical mills that aim to 'Do" turning operations for the part sizes you are talking about here.

3. Different types of grinding (to finish and fit to size) come to mind (for harder materials) … Particular jig grinding... With a Moore (tool company) level spin table and precision sine plate that external radii of interrupted cuts and even external short tapers can be brought to high tolerances.

4. Hand work / fixes / artful design and builds and various hacks... WHY 'cuz not everything can be delivered by CNC milling or turning. For example with air spindle can you can execute some grinding operations on a Hermle but "They" don't recommend it as it will bugger the machine (ultimately)… Other machines can be hacked into bastard jig grinders. There may be other techniques that employ hand lapping of certain surface or various kinds of lapping jigs that could be built to make your high tolerance components fit ----> For example if you are making precision assemblies that resemble bearings of one kind or another that both have linear and rotational elements.

5. In other words Kern or Yasda may not be the answer, just a different process or set of processes, depends on application + actual functional surfaces and tolerences and how they reference each other. Set up time and number of precision set ups and actual cycle time are a major consideration IMO and hence a B axis mill turn machine can be very efficient and accurate. Otherwise a part has to walk around the shop between 5 machines and numerous set ups... Lot of fiddly set ups that require a steady hand and Ritalin (they don't go together very well lol), and plenty of opportunities to screw up OR you design elements into your parts for precision fixturing.

With some B axis mill-turn machines in cycle gauging can be applied (using 3d probes like from Renishaw to map out the topography of a gear or fan blade to plan final cuts) … That's a whole different set of rock throwing and set of arguments.

___________________________________________________________________________

@ricardo_gt can you give any idea of form tolerances ? And what needs to relate to what and to what degree (especially angular tolerances.).
 
Use of Wire EDM might be another consideration (for your process) depending your parts but for things that have to interlock or slide past each other very accurately that may be a good goto in combination with lapping etc. (slow cycle times though .).

Random association ---> Yasda , I think of mother machines cutting things like larger cast iron components , machines set on 1 meter think foundations etc. [I like their design philosophy .].

---> Kern , I think of really fine surface finishes for longer cycle time smaller complex components that have the "snot" thermally controlled out of them... Synthetic granite base an good vibration dampening too.

Having a 5 axis machine gyrating for many hours making microscopic chips is not always the answer for functional high tolerance machine parts.

You already have a Hermle so for more 5 axis mold level work (complex compound curved high tolerance surfaces) up from there would be ??? ---> Buy a better Hermle ? [Some of Makino's offerings are pretty damn serious too.].
 
Roughness N1 are all values of Ra from 0.025 or less.
Outside diameter tolerances are 16H6: +0.000 to -0.011 milímeters.
Imagine a square plate with a curved pocket groove were a rounded shape tube will have to fit perfectly and allow slinding of both parts on that surface.
We will fabricate the square plates and the shaped tubes. Can not use lathes because of curved tube shape.
We did the test parts at one hermle. We had a lot of problems achieving the Ra and the h6 tolerance repetibility. That is why we think of a better machine to do that. We need to deliver 2000parts at 2020 and 3000 parts at 2021.
 
Roughness N1 are all values of Ra from 0.025 or less.
Outside diameter tolerances are 16H6: +0.000 to -0.011 milímeters.
Imagine a square plate with a curved pocket groove were a rounded shape tube will have to fit perfectly and allow slinding of both parts on that surface.
We will fabricate the square plates and the shaped tubes. Can not use lathes because of curved tube shape.
We did the test parts at one hermle. We had a lot of problems achieving the Ra and the h6 tolerance repetibility. That is why we think of a better machine to do that. We need to deliver 2000parts at 2020 and 3000 parts at 2021.

Ok cool … Will put thinking cap on and try digest your part scenarios...

Just out of interest what are the rough dimensions of your square plate ?

What Hermle / spec do you have ? (if you don't mind me asking) …

As there are a lot of things that they offer in terms of up-speccing, / higher precision and accuracy models. I'm not doubting your reasoning just need to get a more clear picture of what you are after. That's weird you couldn't hit the 16H6 tolerance on the Hermle with good repeatability unless you have a model that does not have a spindle chiller etc. especially effecting your "curvy" tube ? Not questioning your tooling , tooling choices , tool balancing, tool setting and cutting strategies. ).

Do you HAVE to spend $1M ?

Is that a requirement ?

5 axis mold machine or 3 axis mold machine ? (For your parts ?) Do you think ? Does the curvy tube require 5 axis ?

_______________________________________________________________________________________________


* Not to pry too much I can't help but feel IF the two parts slide past each other that there isn't a simple geometric hack to cut that negative tube shape precisely with a tilted plane and an appropriate cutter ? But maybe the 'Tube" shape has special features where circular/ spherical (sections/ segments/ "Partials' ), elliptical or parabolic or hyperbolic cuts are not possible ? I'm not doubting your grasp of what you are doing just sometimes there are simple yet counter intuitive cuts and crafty set ups that can be made very precisely (with good surface finishes) before one goes to a $1M micro-machining capable mold machine ? [Just trying to rule that out / eliminate/ due diligence to make sure there are no blindingly simple solutions (that are not obvious) that are "Easy when you know how" (kind of thing) [Old school.].].

Also wondering if the parts could be lapped to each other to slide past each other with a good fit / high tolerance contact / bearing surfaces ? There can be some hacks for that (depending on geometry) that can even be carried out on a Bridgeport + reciprocating motion or epicyclic/ random motion + recirculating diamond slurry + "wet box" / sump. Kinda messy but can actually achieve higher tolerances and better contacting low friction bearing surfaces than what a $1M cnc milling machine can achieve. Although the Kern cutting the Chicago 'bean" is impressive never the less. (spherical and cylindrical surfaces / movements can be lapped thing like tapered geometries can't be lapped as random movement perpendicular to eachother is preferred).

____________________________________
 
Very bad draw but i think you can get the idea.
Both arcs, from tubes and plates are the same.
Plates have 15 degrees inclination from top to bottom, are not flat at Y axis, if you understan it better like this. But they are flat at X axis, from left to rigth sides.
We think having some machines dedicated to this will be the best. Management agrees to buy them, so, i just need to know if there are some better machines than yasda or kern to do it, by the same or less cost.

IMG-20190804-200301 — imgbb.com
 
Roughness N1 are all values of Ra from 0.025 or less.
Outside diameter tolerances are 16H6: +0.000 to -0.011 milímeters.
Imagine a square plate with a curved pocket groove were a rounded shape tube will have to fit perfectly and allow slinding of both parts on that surface.
We will fabricate the square plates and the shaped tubes. Can not use lathes because of curved tube shape.
We did the test parts at one hermle. We had a lot of problems achieving the Ra and the h6 tolerance repetibility. That is why we think of a better machine to do that. We need to deliver 2000parts at 2020 and 3000 parts at 2021.

AND


I can't exactly say what parts are, sorry.
Stainless steel high temperature resistant.
Tubular and flat shapes with several grooves. Outside and inside threads. Parts must perfectly match each others.
Tubulares around 120mm long and squares around 150*200*45.
Thank you.



Makino D-200 Z or DA-200 Z (I think) should be capable of that.

Fit your part geometries tolerances and surface finishes + production requirement of 2000 to 3000 parts over two years per year.

They have really pretty good automation options that are expandable and expandable for tools. That where you can put the rest of your $1M.

Base price is about $375K small foot print
(If memory serves me right I believe the machine does NOT need a super proper foundation (it might be on 3 points kind of thing).

I'll dig up a couple of links. Good surface finishes definitely mold level work and good geometric tolerances for mechanical parts (theoretically) … There aren't that many of these "running wild" in the USA at the moment.

Metal removal rates aren't too bad...


^^^ This doe not do the machine justice IMO.

D200Z | Makino


30,000 rpm - HSK-E50 (Standard) , seems interesting.

__________________________________________________ ___________________________________


* No affiliation

*** Double triple caveat have not had test parts run on the DA-200 Z nor own one... The Z- knuckle + machine's linear "Moves" are deceptive and asymmetric BUT should fit your part geometries... In other words the nominal travels give the impression of a somwhatlarger work envelope (at least for symmetric parts.).
 
Very bad draw but i think you can get the idea.
Both arcs, from tubes and plates are the same.
Plates have 15 degrees inclination from top to bottom, are not flat at Y axis, if you understan it better like this. But they are flat at X axis, from left to rigth sides.
We think having some machines dedicated to this will be the best. Management agrees to buy them, so, i just need to know if there are some better machines than yasda or kern to do it, by the same or less cost.

IMG-20190804-200301 — imgbb.com

Sorry posts crossed in the 'Ether"...

Hmmm interesting .Thanks for that.

Check out the Makino D200 Z , worth checking out what they claim. Fantastic for a small footprint machine and ridiculously fast moves. I'd be interested to see what you think... There may be few more detailed PDF's on line where at least you don't have to bug a reseller immediately to get deeper information.

The D-200 Z should have excellent thermal management especially for spindle growth... Also some of their surface "cube" tests for surface finishes cutting in different directions are pretty amazing also...

D_cube_transparent_WEB[1].jpg

I don't know much about the Matsuura linear machines but to my eye thermal management would be something to verify/ test as linear drive mechanisms pack out a ton of heat and counter measures for magnetic suck down needs to be implemented also...
 
Sorry posts crossed in the 'Ether"...

Hmmm interesting .Thanks for that.

Check out the Makino D200 Z , worth checking out what they claim. Fantastic for a small footprint machine and ridiculously fast moves. I'd be interested to see what you think... There may be few more detailed PDF's on line where at least you don't have to bug a reseller immediately to get deeper information.

The D-200 Z should have excellent thermal management especially for spindle growth... Also some of their surface "cube" tests for surface finishes cutting in different directions are pretty amazing also...

View attachment 262488

I don't know much about the Matsuura linear machines but to my eye thermal management would be something to verify/ test as linear drive mechanisms pack out a ton of heat and counter measures for magnetic suck down needs to be implemented also...
Do you have makinos at your shop?
Do they achieve high tolerances with high repetibility?
 
Sounds to me like you need to be talking with a few good machine builders and their production engineers who can actually look at the parts/drawings and do a proper study of the process and figure out ROI of different ways to tackle it. Maybe needs some grinding, sometimes some features can be better done on specialty built machines with custom tooling, strict process, inspection and such.
Don't spend any $ until that's properly sorted out.
 
Do you have makinos at your shop?
Do they achieve high tolerances with high repetibility?

Not for 5 axis, out of my league / price point for 5 axis verticals or horizontals.

The thing with Makino is their 5 axis machines are really fast AND accurate as a production machine. That's why they are normally quite pricey... I think D-200Z at base price at $375 is the "Cheapest" dedicated 5 axis machine they have.

For me Makino are the "Goto" in the machine tool industry as a reference point for what is possible and practical in a quasi production environment for high tolerance stuff. At least for the type of sand pit we lurk in / aspire to be. Meanwhile we fix sh*t in other ways lol. So 5 axis mill turn is a very good fit for what we have to roll out and we can hit our mechanical tolerances well if we design for that method/ platform... It's just the function of our critical parts and part lengths. For future automation + higher demand, larger bore B axis mill turn with gantry loader and bar feeder is waaay more efficient than anything else I can think of for what we need.

They don't make a "slower" entry level makino… I'm not so much of a fan of rotary trunnions mounted on 3 axis machines even if its a mold machine like a makino F3 or F5. Even then the F3 and F5's are made in Singapore (rather than Japan , nothing wrong with that) and Makino is not without reported spindle issues either (tool bending moment not withstanding). Even amongst the best MTBs temporary weak spots can occur in a line, wrinkles that need to be ironed out.

Hermle offer entry level machines like the C-250 etc. but they don't have a spindle chiller or may be missing one rotary encoder on an axis etc. but something like a blinged out C-22 with their high precision options (as a specific enhanced demanded option) , tandem drive DD drives etc. should go toe to toe with a Makino for form tolerance and surface finish IMO (especially with synthetic granite castings ) but not necessarily for super high speed production. Hermle's HS flex system for automation would hit your 2000 to eventually 3000 finished parts / year (depending on cycle time). It has expandable add on bays and number of tools as you go from 2000 to 3000 parts / year. That's why I kind of think the Makino D-200Z seems quite neat... (doesn't work for our part lengths (right now)) but a lot cheaper than a blinged out Hermle C-22 and is very expandable also.


Mitsui seiki Waaaaaaaaay out of my league and they are specialized machines also (very excellent) titanium / aerospace type applications. More $ also.

@ricardo your key problem as I see it is automation and production numbers + actual part cycle time. That's where your thermal management comes in, hitting your tolerances in an integrated automated cell in a reliable way... While staying under $1M.

___________________________________________________________________________



Yasda and Kern are very specialized and have their own strengths. Not sure you are going to have much spare change $ and cents to automate a Yasda or a Kern from $1M ? [I have old pricing on Yasda but never pursued Kern … (too specialized for what we have to do right now.).
 
Sounds to me like you need to be talking with a few good machine builders and their production engineers who can actually look at the parts/drawings and do a proper study of the process and figure out ROI of different ways to tackle it. Maybe needs some grinding, sometimes some features can be better done on specialty built machines with custom tooling, strict process, inspection and such.
Don't spend any $ until that's properly sorted out.

Hard to tell from the picture but I was kinda thinking form grinding also.

Definitely agree without detailed specs, drawings, models and goals … hard to "Be " definitive.

_______________________________________________________________________________________________________________


The old school Die hard part of me can't get the idea of form tools and tilted plane on a precision rotary / spin fixture...

Just for curiosities' sake more precise application information is not available, ~Yes it's stainless (in some sort of state ?) but Ra 0.025 … ?

I wish OP would use proper units Are we really talking 0.025 micron ? or 25 micron (confusing with the diagram on the pad (chunky spirals)) ?

That's a 1-micro inch finish … Not a 16 or or 32 but 1 micro inch Ra.

https://www.mscdirect.com/product/details/06558159

Sample ^^^ surface finish comparator.


I'm not 100% clear on the real form tolerances here either ?

0.025 micron, 25 nm (nano-meters.). 20 times smaller than the wavelength of green-ish light. Better than optical molds etc. as milled.
 
Last edited:








 
Back
Top