What would it take to use ONLY CATIA for gouge detection? Please do NOT discuss Machine sim.

Hi Friends,

(1) What would it take to use ONLY CATIA for gouge detection?  Please do NOT discuss  Machine sim... I want to be clear... I am only talking about finding gouges, holder rubs, fast feeds, excess material for roughing.... etc.

(2) As a side question... how important to you is the CATV5 interface.   Again... Regular workpice view.... NO MACHING SIMULATION... .I only want to compare Vericut to the CATIA Gouge detection.  

(3) A related topoic while talking about the interface ..... Would the lack of a CATV interface for V6, slow you down in any way, if you were close to move to upgrade to V6...... whould that cause you some hesitation in moving to V6?  On a scale from one to ten..... how important is the interface to you?

(4) ..... but yes...... G-code is relevant to this topic. Why.... because verification involves the option to verify got the center line file and also MCD. Yep... no Vericut... that means you need Control Emulators.

(5) Finally.... are you aware of the latest evolutions in CATIA Verification? It has improved. Do you know what is new? Would you like a Ask The Expert to see what is new?

This is important so if you are one of the guys that is sitting on the fence..... please get off your butt, and speak up!!!!!!!!!

Thanks in Advance

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

BIG PROBLEM

To not have CATV work with V6 would be an complete and absolute show stopper for us.

No IFS ANDS or BUTS about it.

If there was no CATV, I think we would seriously have to consider whether to go with V6 or some other CAM system when we reach the time to migrate.

Randy Hitzeman

That's a great question, Dave.

Catia does show some false collisions sometimes.  Especially with axial entry or exit near the part with Roughing MO's.  No big deal.  I usually ignore them.  They can be removed by setting video replay settings very fine, but the performance loss isn't worth it to me.

Other than that, I do 100% trust Catia for fixture/part/stock/shank collisions.  We're talking COLLISIONS here.  Not gouge/excess detection.

For that, Catia isn't worth a crap either with or without machine sim.  All you get is a bunch of colors. Red or Green or Blue.  No way that is good enough to run an expensive part.  The collision/gouge report they give you for free only works with 3-axis toolpaths.  If they made it work with ALL MO styles and tool axes, then that would be getting close...

What would it take?  A complete gouge/excess report stating which MO and the XYZIJK point-vector where the error occurs.  No machine sim is neccesary for this, but it should work with or without machine sim, because you should be able to have these values WITHOUT a kinematic machine, and running just the MO's or imported aptsource file.

Basically, if they add this ability to their "analyze" function (I don't know if all you guys even have that.  It costs extra, but came with my PLM express configuration, which is one of the main reasons I went with PLM express- that and Multi-pocket, and Prismatic asst, etc...).  You click analyze, and the cut stock turns all different colors based on whether there is a normal, excess, or gouge condition.  It works well with the right settings, even on big parts, but not on huge, complex parts- it just takes forever...It needs a report function.  Then, I would say that it would be possible to ONLY use Catia for gouge/excess/collision detection.

There are tricks to use Catia....use the design model for stock, and shrink or expand your cutters....stuff like that, and set it to stop at any touch.  But this is not good enough to sell.

Just about everything CATV does can be VBScripted, so maybe that is a possibility.  Someone will probably write up the script and sell it....in fact some shops use their own script to do the same basic thing...rather than use CATV.

In Reply to Randy Hitzeman:

BIG PROBLEM

To not have CATV work with V6 would be an complete and absolute show stopper for us.

No IFS ANDS or BUTS about it.

If there was no CATV, I think we would seriously have to consider whether to go with V6 or some other CAM system when we reach the time to migrate.

 

Randy Hitzeman

WOW Randy,

I will be honest..... I expected some response similar...... as I know the industry.   But for you.... what is missing that makes Vericut so important?

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

In Reply to Bryan Carpio Felsher:

That's a great question, Dave.

Catia does show some false collisions sometimes.  Especially with axial entry or exit near the part with Roughing MO's.  No big deal.  I usually ignore them.  They can be removed by setting video replay settings very fine, but the performance loss isn't worth it to me.

I reported some issues where on a small 5-axis simple part.... like a pliiow with 4 pockets that there were false gouges because of the inacuracy in the graphics.  .002 may not sound like much, but the back and forth 5-axis motion looked like there were mismatches. I thought I might have made a programming error. Then I check it in Brand X, and the surface looks and measures smooth as a babys butt.  If I had not had the Vericut.... I would have been chasing my tail looking for problems that were not there. ..Oh yeah... there was a .0015 gouge that was missed with the colors, and analsys.

Other than that, I do 100% trust Catia for fixture/part/stock/shank collisions.  We're talking COLLISIONS here.  Not gouge/excess detection.

THANK YOU FOR SAYING THE DIFFERENCE BETWEEN COLISSIONS AND GOUGES. WE NEED TO HAVE DS LOOK AT THIS AS DIFFERENT PRODUCTS........ TO GET WHAT I CONSIDER NEEDED REQUIREMENTS...... STAY TUNED.

For that, Catia isn't worth a crap either with or without machine sim.  All you get is a bunch of colors. Red or Green or Blue.  No way that is good enough to run an expensive part.  The collision/gouge report they give you for free only works with 3-axis toolpaths.  If they made it work with ALL MO styles and tool axes, then that would be getting close...

I agree 100%, this is one of my requirements. I have stated many times .....if the colision detection in replay was in both Video, and MS, we are getting closer. It is all about checking what you created on the hard drive, not interactive. Why?, because in CATIA, if you jiggle here, something can move over there. You change some geo for he 150TH MO, that can effect the 34th MO.  ( Note They call it colision detection in replay, but it is the function that look at the gouges

What would it take?  A complete gouge/excess report stating which MO and the XYZIJK point-vector where the error occurs.  No machine sim is neccesary for this, but it should work with or without machine sim, because you should be able to have these values WITHOUT a kinematic machine, and running just the MO's or imported aptsource file.

Very true.... But ... you need a Machine and Control for for g-code gouge detection... without using MS... in Veriuct... Same with CATIA. You need a CE.  Vericut at least give you the basic CE and machine files, without a  MS license. That means the regular product CAN simulate g-code, with no MS license. You just can not simulate the machines.

Basically, if they add this ability to their "analyze" function (I don't know if all you guys even have that.  It costs extra, but came with my PLM express configuration, which is one of the main reasons I went with PLM express- that and Multi-pocket, and Prismatic asst, etc...).  You click analyze, and the cut stock turns all different colors based on whether there is a normal, excess, or gouge condition.  It works well with the right settings, even on big parts, but not on huge, complex parts- it just takes forever...It needs a report function.  Then, I would say that it would be possible to ONLY use Catia for gouge/excess/collision detection.

As I said 4 years ago, I could not get it to work on a large bulkhead. It took like 7 min to compute. Most of the time it crashed. Sometime it would compute, but would crash after rotatiing.

There are tricks to use Catia....use the design model for stock, and shrink or expand your cutters....stuff like that, and set it to stop at any touch.  But this is not good enough to sell.

Just about everything CATV does can be VBScripted, so maybe that is a possibility.  Someone will probably write up the script and sell it....in fact some shops use their own script to do the same basic thing...rather than use CATV.

There is no way for you to know this so I will tell you .... VB is different in V6. One reason is because everything is an object, so VB code that points for directorys and files do not work.  THE BOTTOM LINE:   There is NO WAY to write out a STL or STEP using VB.  The options are confusing to me. My coworkers are sending the assys down to V5, and then running the CATV5 interface, and other alternatives are usign CAA to script the export of STL's .   All these work arounds are labor intensive and we ....... really need the functionality now.  I am thinking .... make Catpart from Catproduct.... or step.... still looking at options. ....

 Dave

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

'Gouge Avoidance' inside the tool path computations for all MOs.

EDIT: Removed original message.

Regards,
Samarinder Singh

-------- Original message --------
-----------

This would not be a good thing. I rely on Vericut to fine tune the program.

The thing is- nearly ALL CAM SYSTEMS have always had "collision" detection.  It is the actual "GOUGE" detection that sets Vericut apart from these CAM systems, and even other CAV systems.  (Hey- did I invent a new word- Computer Aided Verification?).

A while back I looked into buying Predator software.  It was only about $11000 at the time, and came with full machine simulation, machine/CE builders, everything you need to go.  BUT, it only really offered COLLISION testing, because the GOUGE/EXCESS test was only like in Catia.  A bunch of colors on screen representing the part condition.  I told the president of the company face to face and shook his hand that if they added a real REPORT that I would buy it immediately, and that was a promise.  I also told him I could probably sell a crapload of seats.  This was at least 3 or 4 years ago, and I haven't heard back. 

This CAN'T be that hard to program!  A GOUGE is really no different than a COLLISION, other than ANY motion is a violation if it touches the design model!  Any time this happens, just record it, and at the end make a report!  I'm fine with just colors for the EXCESS report, but for GOUGE, we need something in writing- like a .csv file of some sort.

I, personally, think that V6 is the biggest mistake DS could have ever conceived.  Just....plain...a....bad....idea.  When the time comes that I can't make a living anymore using V5 (which I think will be many, many, many years...), I will probably either retire, get a real job, or switch to a new CAM system.  Probably NCL, because somehow, NCL has gotten away without offering any real major enhancement for 20 years.  Some eye-candy stuff, but the basic language isn't much changed.  I like that.  It means the concept works.  Hopefully, by that time, I can afford the fact that NCL takes quite a bit longer to do stuff...By the way, NO CATV in NCL....

Dave, I have a hard time believing there is no way to script writing out a STL or STEP model using VB in V6.  That means there's no way to write doing a hell of a lot of stuff....V6=BAD.  Very BAD, if this is the case, and I think a lot of shops are going to be going in a completely different direction in some years.  Remember, a machine shop/contract programmer only needs to be able to READ OEM models.  They can program their machines using whatever CAM system they please...

Dave asks: What is missing that makes Vericut so important?

It's not what's missing in Catia, it’s what we have invested in Vericut.

 1) We have significant monetary investment in software, time and training.

 2) Our machines are modeled, cutter catalogs created, and an extensive backlog of legacy programs that are already set up for Vericut.

 3) Even if Catia was better than Vericut, which it is not, it would have to be significantly better by a large margin in order for us to even contemplate abandoning Vericut.

 4) Catia isn't the only CAM system we deploy. We also have MasterCAM X5, UG NX 7.5, and legacy programs in GibbsCAM, NCL and Catia V4. The common thread in our programming department is Vericut. 

 5) We occasionally offload programming in peak times. We would prefer, but don't require that our programming vendors use V5. What we do require is that they have Vericut, and use our machine files. CAM systems may vary, but Vericut is universal in the industry. We review the incoming Vericut files as part of our program acceptance criteria.

6) We utilize OPTIPATH in Vericut and are happy with the results. Because it is platform independent, it doesn't matter the source of the program.

7) We find machine control building and customization in Vericut to be very user friendly. From what I've seen, Catia is more of a "Black Box" solution supplied by 3rd party vendors. 

If Catia improved their simulation would I use it? Yes, but it wouldn't replace Vericut.

Because Vericut is so central to how we do business, if we could no longer interface Catia to Vericut?

Well, let’s just say we have more CAM options available to us than Verification options. 

Randy Hitzeman

Magellan is looking to use one CAM system across the divisions ( so I hear ).

Not having access to Vericut would be a show stopper and open the door for another CAM system.

I agree with both Randy and Brian , they have made valid points.

I fully admit that I do not fully understand what's happening here............but here's what I'm getting out of it.

(And I'm not too sure I like it too much)

I understand the whole V6 thing is where Dassault needs to go.......ie: the big global collaborative customers.

It's a HUGE undertaking............and a GIGANTIC leap of faith for ANY company at this point in time, in my opinion, to jump in.

The thing I have a hard time grasping is how this super-large-scale-do-it-all-PLM-system thing is going to affect the downstream

vendors, like the machine shops etc .

We are a subcontractor, we make parts, we do NOT globally-concurrently-design parts......

We need the OUTPUT from a designer for our business to start.

We use Vericut, hell EVERYBODY uses Vericut in machining.

I think we will continue to use Vericut for many many years, AND WE WILL BE ABLE TO EVEN IF WE DO GO TO V6.

It's just that in the future, the DS platform is not going to make it any easier to do so.

The CatV interface has many workarounds you can do and still get to Vericut by making your own .stl files etc etc.

It's just that DS is not going to spend a dime to allow you 'easier' access to a software package that is not a business partner of theirs.

Obviously (and this makes good business sense) they want to steer the Manufacturing community to V6, and the entire Delmia suite.

The question here is how many users is this going to drive away..........or how many users is this going to force to stay frozen at V5.

I'm assuming that DS is trying to ascertain this as we speak.

Yet ANOTHER great reason to show up at the COE conference and be heard.

I purposely am not trying to compare the virtues of the Delmia Product and Vericut for gouge checking as you can read by many

people who have already replied...........the Delmia Product is NOT AS MATURE as the Vericut product is today. Case closed.

In time...............who knows.................let's see what evolves.

Thanks.........Steve 

Hi,

Almost 2 years ago, I purchased Vericut(with all bells and whistles including option to read V5 catparts) for the company I work at. I could have got 2 or maybe 3 licenses for Catia machine-sim for the same amount.

Catia offers a very nice gouge detection and there is no doubt about it. In fact, I have seen its accuracy where UG failed to report the gouge on its own sequential-mill. I found it only after I imported aptsource inside Catia and then checked it again with Catia's collision-check-replay. I found .007" gouge in walls.

But all that accuracy stays inside of Catia. Toolpath gets converted to aptsource and I have seen issues in the past for the circle records. Then comes the post-processor and the toolpath is processed again. Then what happens to the part with G-code, that's what counts. And Vericut does an excellent job for doing gouge-check analysis. I believe it is second to none.

For example, I was machining a part with a very tight tolerance. Imagine you are driving a full 5-axis motion with the 90-deg attachment and you want to make sure that surface produced by the tool-motion is well within 0.003" mismatch tolerance. As far as I know there is no way of checking it inside Catia because they rely on the third party G-code controllers.

If V6 doesn't allow you to save the stl files then it is not a big problem for the companies (who have optional license in Vericut to read catparts directly) as long as V6 will save V5 files. But it would be a showstopper for the folks who rely on stl files.

This is all great feedback gentlemen. I see some common points. And we still have not heard from a lot of you.

 Please take the time to advocate for you own companies interests.

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

OK, Here is Dave’s ideas for Gouge detections Requirements. Most are not new, and have been discussed here before. Some are specific…… but I try to …. State the requirement, without stating mandatory implementation.

1. Check for gouges of the part with absolute certainty.
2. Have a user definable mechanism for filtering out small gouge readings, caused by inaccuracies in tessellation
3. For roughing, mechanism to Check for a minimum desired excess material with absolute certainty.
4. Have a comprehensive gouge list.
5. A methodology or possible a new workbench, for supervisors to verify the final saved program, both using aptsource and c-code. We can assume MS and CE will be purchased.
6. A methodology inspecting… probing the in process models
7. A methodology for creating MBD FTA for the in  process models
8. With a CE, doing just gouge detection with g-code, and without bringing in the full machine, and checking out a MS license is part of the functionality.
9. A methodology for converting the in process models, into catparts.  ( an industry requirement.. Parker Hannifin has been looking for this functionality for years)
10. Some automation to add points on all cylinders, (holes) and check from the exact points on the holes to the math model, not the CGR.
11. Be able to see, and measure the volumes of the gouges. Have transparency of the part, while making the Volumes of gouge solid. A list, where you can click on the list and reframe on the gouges.
12. Replay has and exact gouge check. That functionality is a good start if in Video and Machine simulation.
13. DMU space analysis can check the part to the CGR.  It would be better if it checked to the math exact in process model….. and if DMU was included in an Advanced Machining License. To be competitive… all this needs to be within reach of existing small customers.

Please reply and identify which ones are important to you ... and add your own...14,  15,,,16

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

I agree with Randy

1)      The frontend cost would be significant. Not only monetarily but time as well.

2)      Thru  5)  complete Agree

6)   ---

7)  Machine builder, Machine simulation and emulator is NOT user friendly.

Basically if V6 lacks CATV interface we will NOT migrate.

In Reply to Bruce Sheldon:

I agree with Randy

1)      The frontend cost would be significant. Not only monetarily but time as well.

2)      Thru  5)  complete Agree

6)   ---

7)  Machine builder, Machine simulation and emulator is NOT user friendly.

 

Basically if V6 lacks CATV interface we will NOT migrate.

Hi Bruce,

thanks for taking the time to review the major points, and voice your opinion.

Re

Basically if V6 lacks CATV interface we will NOT migrate.

1. I have posted what I see would get CATIA to where is it better

2. I know the industry, and think that even if Dassault invested and made these changes, the real market share for GOUGE DETECTION... is to be a diferentiator bwtween CATIA and NX. The Buisness case for Gouge ER is more competative with and take teh $$ out of UG's hide. Now for MS.... yes as people use CATIA MS,  the $$ to DS for MS will come out of Vericuts hide.  BUT.... many I talk to will allways want an independent check, from a proven, trusted, 3rd party Verification system.

RE: Machine builder, Machine simulation and emulator is NOT user friendly.

My COE Presentation will show that CE for control emulation is not as scary as it used to be,  and I will teach how to get there.  First, the easyiest way to create a CE, is to have a POST in ICAM... then it is a button push to get 95%, and you use the same kind of macros for post building in CE building.

You DO NOT have to start from scratch at all. The ICAM post, ( and probably others) come with a library of machine configurations, and also controls. To start a post, you select a from these 2 pulldowns, and you are for basics... ready to post.... even on a Siemens 840D!!!

For COE year round, my presentation will probably be voice recorded... but I urge you to try to find a way to get to Orlando. Just my presentation will be worth the money.

If you can not get there, since SPIRIT is a COE member... I offer to do a test dry run, as I alway dress rehearse my presentations.

Trust me on this.... there is ROI for CATIA integrated machine simulation, expecially of you have 5-axis, 4-axis, lathes, and Mill Turns.

Thanks again for posting.

Dave 

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

In my current position, I don't get to use Vericut anymore. However, I don't program any real expensive or complex parts anymore so I get by. From what I see with CATIA, they aren't at the same level as Vericut at this time. And, besides collision detection, our NC guys (the ones that make real programs) use the OptiPath module to improve machine time - they seem pretty happy with that.

In regards to the CATIA NC Full PLM issue, I have been living in the ENOVIA world for 6 years or so and so far, I've not had anyone that could tell me how to fully leverage the power of ENOVIA with the NC package. Maybe V6 takes care of this, I don't really have much knowledge of where V6 is going in that respect. However, if they don't treat all of the different domains the same in the context of PLM, then what's the point...

Just my 2¢

In Reply to Larry Crano:

In my current position, I don't get to use Vericut anymore. However, I don't program any real expensive or complex parts anymore so I get by. From what I see with CATIA, they aren't at the same level as Vericut at this time.

Most users say they use CATIA first, and then Vericut. We all have made suggestions to make CATIA better. We can hope the future holds some new functions.

 

In regards to the CATIA NC Full PLM issue, I have been living in the ENOVIA world for 6 years or so and so far, I've not had anyone that could tell me how to fully leverage the power of ENOVIA with the NC package. Maybe V6 takes care of this, I don't really have much knowledge of where V6 is going in that respect. However, if they don't treat all of the different domains the same in the context of PLM, then what's the point...

Now that I am at Bell, I hope to do a presentation next year on "Why V6 is good for NC programming", since we are on V6. I am a little too green on it so far.

 

Dave

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

In Reply to Randy Hitzeman:

Dave asks: What is missing that makes Vericut so important?

It's not what's missing in Catia, it’s what we have invested in Vericut.

 1) We have significant monetary investment in software, time and training.

 2) Our machines are modeled, cutter catalogs created, and an extensive backlog of legacy programs that are already set up for Vericut.

 3) Even if Catia was better than Vericut, which it is not, it would have to be significantly better by a large margin in order for us to even contemplate abandoning Vericut.

 4) Catia isn't the only CAM system we deploy. We also have MasterCAM X5, UG NX 7.5, and legacy programs in GibbsCAM, NCL and Catia V4. The common thread in our programming department is Vericut. 

 5) We occasionally offload programming in peak times. We would prefer, but don't require that our programming vendors use V5. What we do require is that they have Vericut, and use our machine files. CAM systems may vary, but Vericut is universal in the industry. We review the incoming Vericut files as part of our program acceptance criteria.

6) We utilize OPTIPATH in Vericut and are happy with the results. Because it is platform independent, it doesn't matter the source of the program.

7) We find machine control building and customization in Vericut to be very user friendly. From what I've seen, Catia is more of a "Black Box" solution supplied by 3rd party vendors. 

If Catia improved their simulation would I use it? Yes, but it wouldn't replace Vericut.

Because Vericut is so central to how we do business, if we could no longer interface Catia to Vericut?

Well, let’s just say we have more CAM options available to us than Verification options. 

 

Randy Hitzeman

Hi Randy,

I now have 2013X, and can say, that there is still not way to use CATIA Version 6, with the CATIA to Vericut interface.

You could purchace a CAA license, and I beleive with that, you could write your own interface to Vericut... but your money would probably be better spent elsewhere. That is up to each user.

Perhaps a future version of V6 will has the same "stuff V5 has" to allow the 3rd party softwares, that your company is partnered with to link up with CATIA, ICAM or whatever. 

CATIA has made many improvements in the verification modules in V6, but for the reasons you state, I expect people will want to still use Vericut as at least an "independent final Double check"

Dave

Dave Frank  *    Bell Helicopter *  Grand Prairie Texas

Advanced Computer Aided Manufacturing Systems Engineer

COE Product Co-Chairman, Digital Numerical Control

Let me correct what I said... there is a way.  What we do here is export the product out of V6, bring it into V5...... and then run the Vericut CATV5 interface.   This is a time consuming procedure, at a point where ....time is of the essence... and evaluations are always being made. 

Dave