One of the things we are doing as part of moving to the cell system is changing our tool numbers to 8 digit.  We only have four digit numbers right now and with 8 digits it seems that a lot of additional information could be stored in the tool's number.  Say a tool has a number of 12345678.  1234 would be the functional tool number, 8 would be the individual tool number and that leaves digits 5, 6, and 7 to carry information such as if 5=1 it would ALWAYS break check this tool, etc...  Just some crazy ideas and I was curious if anyone else has done anything like this.  We also have 740 tools in our three libraries (Mastercam, Vericut, and tool storage system) and changing all of these will be a HUGE task.  Any thoughts on that?  I'd be interested to hear how other shops went about a transition like this though I assume it doesn't happen very often.

The main reason for the change is the cell uses FNT (functional tool number) and ITN (individual tool number).  BOTH need to be globally unique numbers so if I had a tool that was currently T0063 and I had three of them they would now be ITN 00630001, 00630002, and 00630003.  I guess the FTN could still be 0063.  The idea is to make it intuitive so operators can easily correlate the ITN and FTN.  It is all a little confusing and I'd love to hear some input on the subject.

Bobw,when you rough with arcs filtered toolpaths does the sgi preform correctly,was talking to someone and they told me that when you use arcs that the control processor needs to calculate more and this slows down the sgi performance,do you arc filter the toolpaths and post the out as point to point with your post

I don't really notice but posting roughing point to point makes the programs huge which slows down Mastercam verify and Vericut.  That is the main reason I rough with arcs.  Also, the moves need to be really small or the servos will make some noise and the machine will not run smoothly, at least that is my experience.  I am running A51nx HMCs.  For finishing you absolutely should run point to point.

Does the machine have the high speed machining option?  No toolpath will give good results if it doesn't.

What is the nature of the surface?  Is it a high quality surface, STEP,  or junk STL?  Haas generally likes filtered toolpaths with arcs so I would use a flowline without spiral.  By adjusting the filter settings you should be able to get 99% arcs and the machine will run smoothly with good results.

I need to clarify that I only run point to point when finishing, not roughing.  When roughing I run filtered toolpaths with arcs on.

It sounds like you have the Pro5 control.  Here are the answers to the best of my ability.

1.  Correct, M250 is standard mode, M251 is roughing and M252 is high accuracy I believe.  I'm not sure on M252 because standard mode has always been good enough for what we do.  I modified my post so these can be activated by misc integers and I believe the post can be modified to do it automatically.  These can also be changed mid-program (mid tool path) in the function menu if the machine isn't running smoothly.  Sometimes it gets going so fast I fear it might shake itself apart so I change the mode.

2.   Makinos with SGI.4 run better point to point with small moves.  I typically set my tolerance to .0002", turn on smoothing, present arcs as lines, and set the move length to .005-.010 depending on geometry.

3.  No issues jumped out at me.

4.  I assume you have the data server?  If so you can run from DNC with standard names (widget.NC, etc...).  Any programs called via M198 need to be in O0001 format.

Have you asked Foghorn Leghorn (James) about this one?  He is a master of all things Fanuc, Matsuura, and 5-axis so he should be able to give you some insight.  I don't believe he frequents this forum anymore but you might try sending him a PM.

What make of machine?  Also, why do you want to do it in the tool change macro?  You'll end up checking breakage on all types of tools that don't necessarily need to be checked because there is no chance of them breaking (chamfer mills, spot drills, reamers, etc...).  I built these routines into the misc integers of Mastercam so I can specify them for the individual operations and it has worked great.

Anyways, have you tried running the tool change macro in single block and checked exactly how the machine is reacting as it is executed, line by line?  I'm no expert in this area but that is what I would try first.

Bob, you say (and I believe you) that the Makino produces $200k more in a year then your Haas. I'm curious, did you have a customer lined up telling you you're getting this ramp up in business and you're forced to find a way to get it done? It's one thing to add capacity but it's another to find work to feed that's capacity. It sucks adding capacity and then waiting for that forecast that doesn't come to fruition.

I didn't have any work lined up for the first A51 HMC, it was a complete leap of faith.  I saw it as a way to put the shop in a whole different category quality wise and I also felt it would really make us more efficient in project change overs.  It exceeded my expectations and put the shop on a whole different trajectory all together.  It made the $5 (commodity) part extremely profitable and gave us a ton of versatility in what work we decide to do.  We can do 1000 basic widgets that anyone can do, then turn around and do orthopedic molds that are extremely demanding, and knock them both out of the park. 

Thanks.  I read through Colin's pinned post and figured out how to import the settings.  I ended up copying all three files with new names, selected the copied post in the new control definition, then imported the settings from the old control definition into the new one.  Seems to have the correct settings so far.

Yeah, I know.  I should know how to do it but I only need to every blue moon.  I recently took delivery on a Makino A61nx HMC that will be part of a cell system.  I already have an A51nx machine definition, control definition, and these are dialed but since the A61 will be in a FMS there are a number of tweaks that need to be done to the post.  I would like to copy everything from my A51nx setup to the new setup as a baseline and start from there.  What is the simplest way to do this?

My point is, these differences aren't pocket change, they are serious $$$.  Enough $$$ to make it worth really taking a hard look at more than what the machine costs, but where it puts the shop and what it brings to the table.  My argument above doesn't even mention tolerance or surface finish, those are icing on the cake.  Implementing things like this isn't a cake walk either.  You don't just order one and all the sudden whammo, making 50% more $$$.  It takes a lot of effort and investment to really get it going right and it is a continuous process.

we know you are not just talking, but making it happen.

 For me the things you are often saying seems paradigm shifting from my poor boy mentality of doing things on the cheap as much as possible.

I know the problem with ANY non pallet machine (ie. generic VMC) is that a guy always has his head in them.

 To me, the only way a shop is going to take off like a rocket with your machinery changes is by taking advantage of pallet loading and rotating to different sides, additionally taking advantage of tool magazines. For short job shop runs the speed of the machine seems pretty inconsequential compared to getting jobs on/off machine through tooling and prog methods (which you seem to be killing at) all which can be applied to a VMC to some degree.

 

 It would be an interesting choice between buying TWO pallet Haas' each with at least one nice rotary tombstone verses ONE makino hori. Without a specific part in mind if i had to write the check tomorrow two spindles SEEM greater, IMHO.

Here is what has been driving my philosophy.  Watch one of your machines run and record the actual cycle time of the part.  Now take the value of the part ($100 part with 25 minutes of cutting time for example).  With the value of the part you can determine what shop rate you COULD be working at if you were able to keep the spindles going non stop.  Now at the end of the day have a look at how many parts you actually have finished and sitting on the counter and it will probably represent about 40% of what COULD be done if the spindle was to keep going non-stop.  The other 60% is waste due to the machines, people, and processes that are in place in that shop.  I have made it my mission to find a way to minimize this waste and make it productive, somehow.  The first step was to get horizontal machines.  The second step was to automate as much as possible via macros and technology.  The third step was to standardize our processes, and the fourth was to revisit every program to increase its efficiency and truly utilize the machine's performance.

When I did comparisons between the two machines, Makino and Haas I saw a 25% improvement in speed with the Makino over the Haas with the same program.  Once optimized for the Makino it would have been a much bigger spread.  If I am able to push $75k per month through the Makino with a single shift the Haas could only do $56k.  $56k is waaay too optimistic with a Haas VMC though because it isn't a pallet changing machine.  That is a valid comparison when comparing a Haas HMC to a Makino HMC though.  Two Haas HMCs could outperform one Makino HMC but at the cost of additional manpower and floor space.  My shop had 1200 square feet and given the equipment I have, I may have had the capacity of $1000 per square foot of shop space, just for arguments sake...  If I was running Haas machines it would have been $300 per square foot of shop space.  With Makinos I might have been able to produce $450k per employee where with a Haas it might be closer to $300k per employee.  It goes on and on...  Any way you slice it, it is all about efficiency.  Time, space, materials, etc...

so we are assuming that the Makino will bring in 200K more than the haas a year? In that case its worth a few spindles

When I moved my work from the Haas VMC to the Makino HMC it netted well over that per machine.  I run job shop work, molds, and short run production with pretty loose tolerances.  It was all work that could easily be done on the Haas machines, Makino just did it better, and faster, by a lot.

I keep arguing this because I feel there is the perception that higher end machines are only justified for balls out production, or crazy accurate parts.  This is not the case at all.  They can help shops become extremely efficient and much more profitable.  I know my shop took off like a rocket when moving to the HMCs from Haas machines and became much more profitable.

I have a macro call built into the tool changes of all my machines that does numerous tool checks.  One such check is a minimum tool length check.  If the tool length is shorter than 2.5" it will alarm the machine.  There is a lot that can be done to make things more bullet proof by automating checks such as this, and pallet checks, etc... through macros.

You should never consider a crash? Why has this never happened at your facility? I like the comment, if your operators are that poor its time to replace them... so anyone who crashes a machine is considered a poor operator then and they all need replaced? Sounds like you have some godlike operators over there, I bet they are high paid.

I think what he meant was he wouldn't build a shop and spec machines based on how costly they are to repair when crashed.  Personally, my decision making in this area is based on performance alone.  If the machine generates an extra $200k per year over a Haas, the cost of replacing a $30k spindle when crashed doesn't even figure into the equation.  On my Haas machines I was replacing spindles every two years due to wear, with a day or two of down time each time and that DID figure into the equation of justifying better equipment because it was always at the worst possible time.  Of course all operators have issues, including myself, but if I were buying a new machine and I was really concerned that Joe would probably crash it here and there, The machine would be ordered and Joe would be out the door.  If I thought Joe might crash it sometime in the next 5-10 years, well that is pretty normal in this shop so Joe's job would be safe.

I think the standard/ generic post is set up assuming the B and C axis intersect each other.  Are you doing full five axis machining, or just positioning (3+2)?  If full simultaneous 5-axis I believe you will need to use WSEC and TCP together and it will account for the offset perfectly.  This is no simple task however and requires a post and machine control definition definition set up for it.  If you are running positioning only there are other ways to do it.  You could write a macro that calculated the positional changes using Euler's angles and moves the offset (G54) using G10 commands, no simple task either, but doable.  This macro would be executed at every index.

The easiest way to get everything working perfectly without an act of congress is to adjust the position of the C-axis so it is intersected by the B-axis.  This will take some time and you should shoot for .0005" or better, ideally .0002" or so.  Your best accuracy limit will be double whatever this offset is.

I haven't looked at the file, still on X8, but are the vector curves chained the correct (consistent) direction?

Does it come with the PH10M/10MQ motorized probe head?

Yes, PH10MQ.

This may sound like a silly question but when running a cell of horizontals do you take into consideration the materials you're cutting in each station? I have 5 haas verticals and run nickel in 3 of them exclusively, aluminum in one exclusively. I assume to maximize productivity in a cell your recycle bins are mixed with whatever materials are run through in que. My nickel scrap (not so much recently) is valuable and only worth half if it's contaminated with other chips.

I haven't run the cell yet but when running the horizontals we try to keep the chips separate but it is secondary.  We will not let the machine sit for 30 minutes while we clean out the chips, etc...  The cost of the machine sitting would be 3X the value of the clean chips.

Calibration is also crucial. Measure the sphere in at least a dozen points. Don´t do like average guys, that pick the min. number of points, 5 I think, and start the measurements. Always invest the right time in the calibration of the stylii.

The calibration routines the machine comes with are fantastic and it takes about 80 hits per probe orientation and when complete it provides a report which includes the degree of error that has been comped out.  There is a matrix where you can select the various angles to calibrate at and there are a few hundred in total so one could easily spend 30+ minutes calibrating a probe for every index combination.  Once the orientations are selected everything is automatic.  It takes several hits on the ball and also takes several hits while sweeping/ scanning the sphere.  I have the SP25 scanning probe along with two styli racks in the tool changer.  It was a demo unit and was very well set up.  The max volumetric accuracy error over the travels was 1 micron during the machine setup and calibration.

Hi Bob,

 

Have you looked into the Verisurf options that are available for CMM Programming? I'm sure Ron can chime in to correct me if I'm wrong, but I thought that Verisurf would allow you to use Mastercam as the programming interface to your CMM. That way you wouldn't have to learn a separate software package just to inspect your parts...

I haven't.  I don't know the first thing about CMMs other than they measure things and can be really accurate.  Is the CMM/ software relationship similar to CNC machine/ CAM software relationship where the CMM is just hardware that can be run by a number of different programming software?  I was under the impression that they were more closely integrated with the programming software (Mcosmos, etc...).

And people keep mentioning datums...  What the hell is a datum?

You must be doing well if you are buying new machines. How long have you been in business Bob?

I have been at it for ten years.

This is how it currently sits.  The cell will be here in a week or two and when the riggers are back in the other machines will be moved into place.  We had too much going on to have them do it when the machine was delivered.  The new machine is on the right (500mm A61nx) and the other two are 400mm A51nx machines.  Ultimately we will be getting rid of the vertical and a horizontal and the cell will go along the back wall.  The new machine will be in service by Wednesday.

I would argue that. All we do are short runs. We can turn our Makino horizontals three times faster than any vertical. Plus, we can revert to a running job WHILE we wait on a first article, while the stand alone vert waits, not earning a cent.

I would agree.  Seems like an application where a FMS would kill it.  Fun the first part, put the pallet on the rack and run another job (or three) while the first article is being done, then make the tool adjustments and pull the job back in, or get it back in the queue when it is ready to go.  100% spindle utilization.