Automatically making finish pass contours on solids

[SnahOlah]

Does anyone have any good ideas for quickly making finish passes (for vertical walls) on a solid model other than just picking edges or edge curves? Also that has lead in and out and tool compensation (G41/G42). Surface rough pocket set to facing with rough turned off is almost perfect, but also almost unusable, as it has no means of controlling the depth so you better pick a tool that reaches the entire distance. Or you can make a planar surface to block it at a certain depth in which case it will finish that surface as well (no checked surfaces), and the subsequent tool with more reach cannot be made to start further down. Also it insists on finishing the outer boundaries of all the flat surface even those with no walls (like the tops of bosses). So for me without overcoming these issues the only way to use it is to convert the toolpath to CAD geo and just apply a 2d contour using incremental 0 depth where you avoid the wasted cuts on outer edges of bosses that have no wall and control the depths for each tool.

[Guest]

There is no Easy button....'

If you don't want to spend any time, try using FBM.......

Good for prismatic work if cycle time isn't an issue...if cycle time matters..dig in an be a programmer

Edited May 30, 2017 by Guest

[Guest]

Dig in and be a programmer. Yep. Couldn't have said it better myself. 

This gig isn't for the guy who doesn't want to do any work. It's for professionals who hone their craft.

 

[#Rekd™]

First off I would post a file and find out the best way to program your parts, the worst that could happen is you learn some new techniques.

 

[crazy^millman]

Reminds me of the manager last year who asked me why I was doing 5 Axis programming. He said you know in 5 years you will be out of a job. There is think tank group working on a program to automatically program all part for all machines. You give the program your file and machine and it does everything for you. I laughed and asked if it was the same group I talked to that wanted me to help them? He looked at me with a look of disbelief. How do you know about that group? I said well we had a conversation at a trade show a couple years back. They could do 1% of parts thrown at them. 14 PHD's had put 20000 man hours into the project at that point over 10 years. There have been a lot of advancements in Feature Based CAM no doubt, but the human creativity and problem solving is still something unique.

To get good program you must do good work. Even the best production lines with robots and such didn't happen on their own. It took someone to make it happen.

[Leon82]

Use a pocket cycle to the depth of the first boss. Then another for the next set. Use your top of stock as the depth of the previous pocket. It won't cut air, but if the depths are close you will get pockets that don't cut much.

Pick your multi pass settings  and save them, they will remain there while you program

[gcode]

"window it and cut it" was part of the first sales  pitch I ever  got for a Cad/Cam package back in the early 90's.

It didn't work then, and sometimes it works now, but it's still the holy grail of Cad/Cam.

They may get there some day, but not in my lifetime.

I think 3D printing and near net 3D printing will eliminate a lot of machinist and programming jobs in the future

What's holding 3D printing back is that they can't "print" metal parts with the structural strengths of cast and forged metals.

They are getting better at it all the time though.

I believe Boeing is now 3D "printing" titanium shapes for aircraft structures that eliminate 70% of the roughing involved in building these structures.

 

[Leon82]

18 minutes ago, C^Millman said:

Reminds me of the manager last year who asked me why I was doing 5 Axis programming. He said you know in 5 years you will be out of a job. There is think tank group working on a program to automatically program all part for all machines. You give the program your file and machine and it does everything for you. I laughed and asked if it was the same group I talked to that wanted me to help them? He looked at me with a look of disbelief. How do you know about that group? I said well we had a conversation at a trade show a couple years back. They could do 1% of parts thrown at them. 14 PHD's had put 20000 man hours into the project at that point over 10 years. There have been a lot of advancements in Feature Based CAM no doubt, but the human creativity and problem solving is still something unique.

To get good program you must do good work. Even the best production lines with robots and such didn't happen on their own. It took someone to make it happen.

I read an article a while back where mastercam was working with a manufacturer to do this with a radiation filter for cancer treatment. Every filter was patient specific, and always different. It was planned to be automatic from and emailed file.  I don't know if it worked out

[Leon82]

4 minutes ago, gcode said:

"window it and cut it" was part of the first sales  pitch I ever  got for a Cad/Cam package back in the early 90's.

It didn't work then, and sometimes it works now, but it's still the holy grail of Cad/Cam.

They may get there some day, but not in my lifetime.

I think 3D printing and near net 3D printing will eliminate a lot of machinist and programming jobs in the future

What's holding 3D printing back is that they can't "print" metal parts with the structural strengths of cast and forged metals.

They are getting better at it all the time though.

I believe Boeing is now 3D "printing" titanium shapes for aircraft structures that eliminate 70% of the roughing involved in building these structures.

 

We had our first print with specs if the part were made with an additive process

[crazy^millman]

7 minutes ago, Leon82 said:

We had our first print with specs if the part were made with an additive process

I have helped with printed parts, EB (Electron Beam Welding) parts for different applications. They are doing some really neat things, but when the machine cost over $1 million and the parts coming off of them have a $XX,XXX or $XXX,XXX build costs just to that point and still require another 12-40 hours of machining not something going to replace everything anytime soon. I have been around 3D printing since I came to California. Very cool tech, but when you have been in well over 500 different machine shops making everything from bone screws to rocket parts. You see what is made day in and day out you realize that castings or that billets and the little bit of machining going into it is still far superior today.

[gcode]

I think "window and cut it" is nearly there with dental technology

My dentist has a little table top 5X mill .

When he's doing a crown , he takes a bunch of scans of your mouth, loads them into a computer, spends 5 to 10 minutes with the Cad/Cam

software, loads a blank and starts the machine.

It buzzes around for 30 to 40 minutes and produces a perfectly fit crown. ..  Its not totally automated, but pretty close.

[mkd]

3 hours ago, gcode said:

I think "window and cut it" is nearly there with dental technology

My dentist has a little table top 5X mill .

When he's doing a crown , he takes a bunch of scans of your mouth, loads them into a computer, spends 5 to 10 minutes with the Cad/Cam

software, loads a blank and starts the machine.

It buzzes around for 30 to 40 minutes and produces a perfectly fit crown. ..  Its not totally automated, but pretty close.

WOW

[Leon82]

3 hours ago, C^Millman said:

I have helped with printed parts, EB (Electron Beam Welding) parts for different applications. They are doing some really neat things, but when the machine cost over $1 million and the parts coming off of them have a $XX,XXX or $XXX,XXX build costs just to that point and still require another 12-40 hours of machining not something going to replace everything anytime soon. I have been around 3D printing since I came to California. Very cool tech, but when you have been in well over 500 different machine shops making everything from bone screws to rocket parts. You see what is made day in and day out you realize that castings or that billets and the little bit of machining going into it is still far superior today.

Yea, its only going to be for the super expensive stuff for now.

Pratt is working with it from what I heard

[SnahOlah]

good advice, fall behind and let companies like first cut remove all your business

[Guest]

8 minutes ago, SnahOlah said:

good advice, fall behind and let companies like first cut remove all your business

We use them regularly....they will never take away our work because they can't do our parts complete....we are constantly having to finish what they can't....they provide a quick turnaround that when we are overbooked, can't handle in our shop....

You might want to learn a little about their business model though.....they have made a VERY, VERY LARGE investment in technology...

In an industry where many small shops want to compete with a 10 year Fadal and reground endmills, they will never compete at that level...

Edited May 30, 2017 by Guest

[SnahOlah]

We also use them a lot, I know the bulk of their entire process and have seen over a hundred resulting parts. They can turn many parts around complete and with extremely fast turn around. in the rapid turn prototype industry I work speed can be everything. You guys may enjoy taking longer to achieve the same result (we are talking about simple finish contours here) but I am trying to automate redundant operations, why not? Is there any reason not to want that?

[Guest]

You might want to look to VB scripting.....

If you have any experience there, check out the sub-forum here.....that "might" have the possibility of getting you closer to what you want...

[SnahOlah]

thank you for taking my request serious (I have been very interested in using visual basic with mastercam as I use it frequently with excel). I only got upset at comments to the effect of "be a programmer" my heart is in this without question. I always insist on top quality results and I believe I can get the same results I would get the long route but with less time. Using the toolpath I mentioned I have already cut the time down significantly on complex models. This technology is getting better (program automation) and we need to stay competitive (it's shocking to new hire machinists when you show what first cutting is turning around in one day). I want to use a program like mastercam to solve these problems as I can have some hybrid of the two, both expediting programming of common redundant tasks, and the ability to bring the full complexity to bear in cases where it doesn't produce the result I will accept. 

[jeff]

6 hours ago, C^Millman said:

He said you know in 5 years you will be out of a job.

I've been hearing that for loner than 5 years, and I laugh whenever I hear it. 

[SnahOlah]

Glenn (post guru) just mentioned that visual basic is only supported by an offline compiler, that's unfortunate.

I have a very good working set of ops for aluminum parts that is very near automated, coupled with my picture frame tabbing method which requires no hand work (dropping the block on the table will even release the part) outside of lightly deburring an edge I can produce a broad range of parts with very fast programming and since the part is just a block throughout the whole process regardless of it's shape fixturing is virtually non existent. Seeing this I do believe that in many cases programming can be made very systematic and these methods are starting to be used heavily in industry for short runs. I believe if you break down a machined part in the right way it is usually mostly a series of the same redundant tasks and I am trying to get away from looking at every part as a completely new task I've never done before. I've been doing this for almost 20 years now after all...

[crazy^millman]

9 minutes ago, SnahOlah said:

Glenn (post guru) just mentioned that visual basic is only supported by an offline compiler, that's unfortunate.

I have a very good working set of ops for aluminum parts that is very near automated, coupled with my picture frame tabbing method which requires no hand work (dropping the block on the table will even release the part) outside of lightly deburring an edge I can produce a broad range of parts with very fast programming and since the part is just a block throughout the whole process regardless of it's shape fixturing is virtually non existent. Seeing this I do believe that in many cases programming can be made very systematic and these methods are starting to be used heavily in industry for short runs. I believe if you break down a machined part in the right way it is usually mostly a series of the same redundant tasks and I am trying to get away from looking at every part as a completely new task I've never done before. I've been doing this for almost 20 years now after all...

Believe it or not most us share that same thought and desire. How do you adapt that process to a Swiss, Lathe or a Mill/Turn? Square pegs in Round holes don't always work. One size fits all might work for what you are doing, but for other areas it is not as easy. Best of luck in your process and please keep us posted when you have that load in a model and out in 2 minutes comes a complete program. We have many companies that will pay millions for that. We have models even on the faster computers take over 20 minutes to just load they are so complex and involved. When parts are families of parts you can dial down some processes, but when they change drastically again one size fits all doesn't apply.

[SnahOlah]

You guys are pretty defensive about this subject I can tell. To be fair to myself I must say you morphed my valid point into an invalid one by adding a lot of extra things I haven't said. All I have described is milling of aluminum parts so far. And I never said anything about it taking two minutes or live tooling lathes or one size fits all solutions. I just want to cut programming time down whenever I can specifically in traditional milling and wanted some feedback on the fastest ways people were using to make finish passes. The surface rough pocket set to facing mode and with roughing turned off would achieve that amazingly well and work in most cases if A. It had depth limit controls (turning on facing removes the ability to limit depth ranges). B. it has the ability to not finish the outer boundary of the flat areas like the 2d pocket. You may want to give this a try and see how well it handles everything outside of those flaws. And seeing that it shows the potential to automate finish passes on walls. I apologize for my own defensiveness but I immediately came under assault by a number of you for daring to wonder these things and in a fairly insulting fashion without you guys knowing my history with machine work. 

[crazy^millman]

55 minutes ago, SnahOlah said:

You guys are pretty defensive about this subject I can tell. To be fair to myself I must say you morphed my valid point into an invalid one by adding a lot of extra things I haven't said. All I have described is milling of aluminum parts so far. And I never said anything about it taking two minutes or live tooling lathes or one size fits all solutions. I just want to cut programming time down whenever I can specifically in traditional milling and wanted some feedback on the fastest ways people were using to make finish passes. The surface rough pocket set to facing mode and with roughing turned off would achieve that amazingly well and work in most cases if A. It had depth limit controls (turning on facing removes the ability to limit depth ranges). B. it has the ability to not finish the outer boundary of the flat areas like the 2d pocket. You may want to give this a try and see how well it handles everything outside of those flaws. And seeing that it shows the potential to automate finish passes on walls. I apologize for my own defensiveness but I immediately came under assault by a number of you for daring to wonder these things and in a fairly insulting fashion without you guys knowing my history with machine work. 

Kind of working both way isn't it? You have no idea about some of our back grounds or maybe how many times someone has come in maybe asking the same exact question. Experience is one thing no one can take away from anyone and we all have different ones. The HST toolpaths do amazing work and you have come to see the limits of the Surface Rough Pocket toolpath. The work in Mastercam is being focused on the HST toolpaths and Surface Rough Pocket from what I can tell is is be lumped into a legacy toolpath, but with your 20 years of experience I am have to assume you were already aware of that. I make my living programming parts and consulting with companies. When someone comes in taunting thoughts like yours I will get several emails from customers asking us why I am not doing it good or as fast yourself. Then once all the facts are laid out and they realize the not orthogonal almost no flat surfaces with crazy surfaces everywhere Ti part that is going from 36 lbs down to 1.2lb that cannot be automatically programmed they get it, but it gets old having to teach people what it really takes someone times to manufacture parts. Programming is part of the process. I apologize if you felt me saying it takes good effort to make good work as coming down on you.

I am being very serious in what I said you figure out a good way I got plenty of companies willing to spend good money to use it. With my 30 years experience 15 of which has been with Mastercam I still haven't figured out ways to accomplish what you are asking automatically. It still requires some effort to make good usable and efficient toolpaths. I am always open to better more productive way to do my job, sorry you can in asking in such a way that I have seen doesn't produce the results you are after. Other methods and techniques do, but after 14 years on this forum I guess it sound like an old mean guy just picking when I am really just trying to sort out facts and help point you in the right direction.

Have you looked at Surface Finish Contour, HST Waterline, Flowline, some of the Advanced 5 Axis Toolpaths in a 3D output. Many ways to get good results with not to much effort, but sometimes I still get the best results from just good old hard work making chains and driving them one by one. Do I hate it every time I have to do it. You bet, but we either do the job making the best program possible using what we got or we keep looking for a better way. Glad you are looking and exploring for better ways that is the heart of what this forum is about.

[SnahOlah]

After the years of using mastercam and the frustrations coupled with it I guess I am still willing to entertain the idea that there is a decent amount of automation possible. I truly don't expect full automation at best I want to find how much I can achieve. I don't feel I addressed the level of experience you guys have only the exact comments made. It was quite an assault for me daring to insinuate there was a solution other then the same old way. Maybe review what was said by me and to me, I started with a request for advice and receive a hail of emotional statements in return. I am starting to learn my lesson and probably should ask my unorthodox questions elsewhere...

[crazy^millman]

9 minutes ago, SnahOlah said:

After the years of using mastercam and the frustrations coupled with it I guess I am still willing to entertain the idea that there is a decent amount of automation possible. I truly don't expect full automation at best I want to find how much I can achieve. I don't feel I addressed the level of experience you guys have only the exact comments made. It was quite an assault for me daring to insinuate there was a solution other then the same old way. Maybe review what was said by me and to me, I started with a request for advice and receive a hail of emotional statements in return. I am starting to learn my lesson and probably should ask my unorthodox questions elsewhere...

I sure hope not and sorry you feel offended. We need more people like yourself on this forum.

I have a saying and hopefully it comes across in a positive light. We can't have keys to the castle till we have waded through the mud. Again hopefully you keep asking good questions and hopefully you can see it is a group that cares to help not hinder anyone. If not I wish you all the best.