Accuracy of tool presetters?

[Bob W.]

I have been machining some molds that require pencil machining with a .015" tapered ball mill and I have been having a heck of a time with mismatch. My machine has a laser tool setter and I know the accuracy suffers when the tools get really small. This has me considering buying a presetter because there is a ton more of this work coming down the line but I was curious what sort of accuracy one would expect from a presetter. I would be looking for +/-.0001". Is this reasonable? I would guess the laser is +/-.0005" on the small tools right now but that is just a guess. I would consider a small tool to be less than .062" in diameter.

[crazy^millman]

Bob, that small best practices IMHO is to test cut areas and adjust for the mismatch. Machining and holding +/-.0001 can be done, but to me is more of an art than thinking I can get it with even the best presetter. I human hair is .003 and time you take dust and other things into account how would you not influence your tools from the presetter to the machine? How you do you match the temperature of the spindle to the presetter? Keeping thermal expansion out of the equation has to come into play and some many other things. Does your machine have glass scales? Have resolution down to the .00005. Had a guy work with me years ago that worked for Moore and he had to hold 80 millionth of stuff from time to time and it has to be done on clear days no rain with perfect humidity in a double temp controlled room. They always took test cuts before finishing anything on a test block. Maybe I am not knowledgeable enough of what can be done with the equipment you have, but even with the best equipment I thinking milling tools at +/-.0001 in and out of the tool change is hard enough much less doing it off the machine and then having it be that close in it. Again I might be wrong, but sounds like a tall order to ask for on such small tooling and would like to see what the more knowledgeable have to say about it.

[Zoffen]

Bob do you run your spindle up to running temp before measuring tools?

 

I have heard of people running the spindle at programmed speed until the spindle normalized and then set them with the laser. This would ensure the tool length is measured at the same conditions which the machine will be cutting at. JM2C.

[Brando]

i did some research on tool setters a few months ago

 

a few makes that come to mind are parlec, elbo controlli and zoller

look for models that clamp the tool in the spindle and have video imaging, those type are within that spec, but they are quite salty

some do an incredibly detailed tool inspection, i was quite surprised at some of the things that can be checked

i would also be looking at does it network well, and is the software user friendly

i'm sure you want to be able to set up a job of tools via the setter , save the lengths\dia offsets to a file\folder via the network so you can just load tools and the offsets file at the machine

[civiceg]

+/- .0002 is what Ive seen from renishaw ts27rs. Ive had more problems with thermal growth between tools, 1st tool running for 2 hrs and warming the machine up completely, then next tool cutting slightly deeper due to being fully warm. This is even with running warmups for a hour plus. Also keeping tool to tool rpm close helps. New Matsuuras and Robodrills, more of an issue on Robos

[Bob W.]

Yeah, thermal growth is a big issue that I have yet to track on either of my Makinos. I know thermal shift is unavoidable but I haven't taken the measurements to know exactly what I am dealing with. Most of the work we do doesn't require tolerances better than .002" which is why I haven't taken the time to determine how much shift there is during the day. Measurements I need to take are coolant temp in the morning and then again in the afternoon after running hard all day. Also probing a reference position in the morning and again in the afternoon after running hard all day. I am curious what probing routines people have developed for their machines to check accuracy. I created a sequence for the Makino with the rotary installed to find the exact locations of the B and C axes in space and it actually brought to light a clamping issue that caused B-axis accuracy issues :-). This was fixed by a few parameter changes as directed by Makino. That probing routine has been very handy indeed.

[GoetzInd]

You can get there with a nice pre-setter WITH power clamping and motorized movement. We have a manual one and you can get under .0005 pretty easily. We have tool probes on all our machines but only use them for breakage.

 

Mike

[YoDoug®]

What kind of Laser? The accuracy varies and the standard units that most builders ship with machines are not the most accurate. It has a lot to do with conical focus points of the laser and noise filtering ability of the electronic hardware.

[Bob W.]

It is a Renishaw laser on a horizontal machine so the head and receiver are about ~30" apart. I know the spacing affects accuracy quite a bit so I know it isn't the most ideal setup.

[Redfire427]

Bob,

 

We have Blum A2 lasers on our V56 and F5 Makino's and measure tools right down to .2mm ( .008" ) holding .0001" accuracy every day. We use a "plunge style" tool setter on the PS95's and S56 and also hold tenths with no issue either. Keep in mind that a proper warm-up is essential and once the spindle becomes "heat-soaked" growth will be minimal. The temperature controls on the V56 and F5 are obviously far superior to that of the other Makino's, but I am surprised you are having difficulty. Our shop is environmentally controlled which also helps and the machines all have proper foundations for maximum stability.

 

When I took my Die-Mold training at Makino 10 years ago, they explained that in order to achieve proper blending, you need to run the spindle at its programmed rpm for ten minutes, then touch the tool off the laser and begin cutting. This is especially important when going from a tool that operates at low rpm ( low spindle heat generation ) vs a tool that runs at high sustained rpm ( high spindle heat generation ). Our higher end Makino's came set-up from the factory to check tool length before, and after, each cutting cycle when you execute a tool change. You could modify the tool change program to add a timed dwell to allow the spindle temperature to stabilize before measuring tools. Keep in mind that when you try to achieve high levels of accuracy like this, it comes at a huge time penalty in regards to production.

 

Do you want it right, or do you want it fast? Two totally different approaches.

 

Carmen

[Bob W.]

Thanks for the info Carmen. I know the accuracy of the laser as quoted by Renishaw is a function of the distance between the laser and the receiver so the accuracy of the unit when they are 6" apart as on a VMC will be 5X the accuracy on a horizontal machine where they are spaced much further. I hand't given much thought to the spindle rpm and the heat issue there. The tool I was having trouble with is in my air spindle so the machine's spindle RPM is zero. It was following a tool that had been running for an hour at 14k. I am generally running these molds overnight so ten minutes here and there isn't an issue at all. The parts need to be right, period. Waiting 10 minutes to let temps stabilize so the blend is right will save a few hours of sanding and blending by hand off the machine. Small price to pay and it would allow me to achieve both goals, right and fast(er) than fixing later :-).

 

Sounds like a good application for setting up a misc integer in the control definition/ post processor. Setting the misc integer to 1 adds a 10 minute dwell before setting the tool.

[Redfire427]

I've used those air spindles many, many years ago on an Okuma I had that only had 4000 rpm. The air spindle gave me 25000 rpm, but I suffered the same issues you are describing. I'm not sure there is a solution because there would be no temperature control on an air spindle. The best you could hope for is to run it at speed and let it stabilize via thermal growth, and then touch it off.

 

I made the assumption you were using the Makino spindle and that is why I was surprised with your issue.

[Fred @ Slate Industries]

+1,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 (LOL) !

"You need to run the spindle at its programmed rpm " till its at thermal expansion

Every Tool Change Along The Way.

There is a parameter" Cycle dispersion" (SP)?

Where if ANYTHING dust, carbon, vapor ,whatever gets sights of the Beam it will Stop the whole show and it won't allow the tool to carry on and in my case Won't Run All Night Unattended.

 

(Blum / Mikron HSM) MOLDS

 

As stated in threads previous

"There is a price" for .0002 / .0001 in headaches but People FREAK when they see the finished product

Electrodes look like "Molded Plastic" Hard milled Steel looks like something you would find on

A U.F.O. on Unsolved Mysteries.

[Bob W.]

Thanks for all of the replies. I'm soaking all of this up like a sponge. Good stuff! :-)

[crazy^millman]

Thanks for all of the replies. I'm soaking all of this up like a sponge. Good stuff! :-)

 

Bob me too.

[Chipmakr]

We have blum lasers on our Mikrons. When I have jobs that really need to blend well with a series of tools I touch them all off @ 3000 rpm & I look at the radial value and make sure I am seeing what I want to see. Then I run a program that calls each tool number up at the speed it will run and dwells for varying amounts of time. For instance my rougher may only dwell for two - four minutes, my semi finisher six minutes, finishing tools for 6 minutes. My experience on a 3 axis VCP600 is that 6 minutes dwell at speed will result in all the tools being "synched up" nicely when they run. I am running spindle speeds in this example at 25,000 ~ 42,000 with an HSK40e using Schaublin D20 collet holders or Schunk tribos holders. I have had similar results on our 5 axis machine, but I will also calibrated the laser with a standard prior to all these steps if I am really chasing perfection. Hiedenhain has a function that allows you to use the handwheel to adjust for good blends as the program is executing. I have never used it but it sounds like its made for just this type of situation, maybe Makino has something similar. I definitely think the air spindle is and additional variable you will have to work out, but if its got repeatability as far as thermal growth goes once its up to speed you should be able to factor it in as well. Running tools in the machine at speed with real world thermal growth & run out conditions would seem to be better than a pre setter IMHO.

[rickcact1]

We have blum lasers on our Mikrons. When I have jobs that really need to blend well with a series of tools I touch them all off @ 3000 rpm & I look at the radial value and make sure I am seeing what I want to see. Then I run a program that calls each tool number up at the speed it will run and dwells for varying amounts of time. For instance my rougher may only dwell for two - four minutes, my semi finisher six minutes, finishing tools for 6 minutes. My experience on a 3 axis VCP600 is that 6 minutes dwell at speed will result in all the tools being "synched up" nicely when they run. I am running spindle speeds in this example at 25,000 ~ 42,000 with an HSK40e using Schaublin D20 collet holders or Schunk tribos holders. I have had similar results on our 5 axis machine, but I will also calibrated the laser with a standard prior to all these steps if I am really chasing perfection. Hiedenhain has a function that allows you to use the handwheel to adjust for good blends as the program is executing. I have never used it but it sounds like its made for just this type of situation, maybe Makino has something similar. I definitely think the air spindle is and additional variable you will have to work out, but if its got repeatability as far as thermal growth goes once its up to speed you should be able to factor it in as well. Running tools in the machine at speed with real world thermal growth & run out conditions would seem to be better than a pre setter IMHO.

 

 

I do the same thing. M300 and M302. M300 is a set time warmup and M302 warms up til the spindle sensors do not sense any heat change for a certain length of time. Then you laser the tool length. Works slick

[BenK]

Thanks for all of the replies. I'm soaking all of this up like a sponge. Good stuff! :-)

 

Me too, this is something I have been trying to get worked out as well. I run a lot unattended and my tolerances are only getting tighter.

[Jim at Gentex]

Yes, this all good stuff.

 

Let's not forget the programming side too.

As a rule, I try to NEVER blend surfaces with different tools unless it is absolutely unavoidable.

 

The more surfaces I can finish cut with the same tool the better.

[Bob W.]

Two pieces of equipment that I am seriously looking at getting are a presetter or a CMM. Based on the advice here my issue isn't necessarily the laser but my methodology and managing thermal changes. Looks like a CMM would give me a better payback and more capabilities at this point. Fantastic info, thanks a ton! Saved me a pile of cash for sure.

[Bob W.]

Yes, this all good stuff.

 

Let's not forget the programming side too.

As a rule, I try to NEVER blend surfaces with different tools unless it is absolutely unavoidable.

 

The more surfaces I can finish cut with the same tool the better.

 

I agree, I machined most of this mold (non-cavity) with a 3/16" end mill to get all of the flat steps perfect (as accurate as the machine) and as a rule I will always do that if there are tolerances or blends involved. In this case it just didn't make sense to machine the entire cavity with a .015" end mill due to time constraints. Maybe next time I'll just bite the bullet and give it a try as a learning experience though.

[Bob W.]

There definitely is an art to this mold stuff.

[Jim at Gentex]

There definitely is an art to this mold stuff.

 

Oh no doubt!

 

I think all 5-axis machining has a component of programming finesse'.

Definitely part science and part feel.

Molds are a particular challenge with their own set of variables.

 

I work mostly with composites and polycarbonates which also have their own unique set of issues.

 

+1,000,000 on the learning experience, Bob.

Whatever I do, whether it comes out the way I wanted or not, I look at as a positive when I walk away from the job having learned something.

[Chipmakr]

+10000

Looks like a CMM would give me a better payback and more capabilities at this point

 

Our CMM is vital to our business. Its allows use to quickly get feedback on all types of geometry. It removes operator influence and once the initial work is done, indepth inspection of many parts is very economical. Being able to inspect at that level revels so many things about your processes. It is a real eye opener. Plus the electronic record it so easily provides. There are just so many upsides to adding this capability. It will certainly help your shop evolve to the next level.

[Sticky]

I run a 15k spindle hmc on a 27 pallet fms, the only way I get good blends on long cycle times/ high rpm programs is by in process tool length measurement.

 

I don't think I'll ever buy a tool presetter for this reason, if you want it dead nuts, you need to measure it in process. Thermal shift is a b1tch. Make sure your spindle chiller is working properly too. The refrigerant in my spindle chiller recently left the building and I had problems on any of the long/high rpm cycles. I have a few different tombstones that have 4-6 hours of time on each of them, some of the tool paths are 1-3hrs per tool and sit at 10-13k for the whole time. I can't get decent blends on those paths unless the spindle is warm and stable. And usually if its fully warm and stable I don't HAVE to measure in process, but I often do on anything critical for peace of mind. Don't be shocked if you see up to .002" of thermal growth from cold to fully warm!