First Horizontal Program

[Sticky]

Sounds to me like you have an issue finding center of rotation. 

 

I program off center and I still need to construct all the same planes. Only difference is I don't set a different offset for each index. Its the same amount of work for the programmer. Difference is you need to make sure all your fixtures are built correctly. Your process is only as good as the work you put into it up front.

 

Would you be willing to share how you came to the conclusion that I don't know how to find cor?

 

I think it's still slightly more work with multiple wcs', getting your post right, building a robust manufacturing process around it etc, but you are right, it's not like it's mountains of more work.

 

 

 

Sticky man you really need to learn how to read. You keep generalizing my comments to mean what you want them to mean and not what they are saying.

 

I have a part right now with 11k surfaces on it. It has about 400 different pockets at different angles on it. If I program it correctly and do my job correctly it like the 100's of others parts the customer has run using one workoffset will come out correctly. Following your mind set it would be 400 different G10 lines for each one. Following my mind set it will be one.

 

You have no clue about my experience or my abilities and not sure why you think attacking or calling it into question is important. You are more than welcome to walk a few weeks in my shoes and get a good idea what I do and do not know. Those who know what and where we are working have confidence in our ability to get the job done.

 

My reading isn't the problem Ron, the issue is much more your lack of interest in using basic grammar in your replies.

 

You clearly stated that the operator finding 28 offsets is waste of time (I agree), which is why I said that isn't the right way to do it, the operator shouldn't have to find any offsets, the post should output them. Then you came back and said you didn't need to find any offsets, which is contradictory of your first statement, and now you are accusing me of cherry picking.

 

My recommendation is to slow down and work on trying to make legible responses so people can actually understand what you are trying to say.

 

As for your 400 offset example (what's wrong with having 400 offsets?), if the part can be made with one offset, fill your boots. There are lots of parts out there you can make with one offset. Is a robust and flexible method, no, but it can absolutely get the job done in many circumstances. My only problem with what you've said so far was claiming that using one offset is good practice.

 

As you can see a couple other guys understand how and why that is, but you have me singled out, why don't you just come out and tell me how much you love me

[BenK]

Would you be willing to share how you came to the conclusion that I don't know how to find cor?

 

I think it's still slightly more work with multiple wcs', getting your post right, building a robust manufacturing process around it etc, but you are right, it's not like it's mountains of more work.

 

If you have a problem on every type of machine you work with than its time to rethink how you are approaching things.

 

If you feel its more work and that makes you feel like the bigger man than by all means go for it. The one thing I have learned over the last few years is there is no reasoning with you on this. In your mind you are the best and everyone else is just stupid. I'm sorry you need do that to validate your self worth but some day you will realize there is more than one way to do things.

 

By the way you don't need to be an @ss to every one on here. It doesn't make you look very professional.  

[Bob W.]

I typically program to COR and when I need tighter tolerances I add a probing routine that will find COR using multiple hits on a tooling ball at multiple indexes and then populate the DFO or WSEC parameters to account for discrepancies.  The probing routine is a parametric macro and is self contained and called as a custom drill toolpath in Mastercam.

[Sticky]

If you have a problem on every type of machine you work with than its time to rethink how you are approaching things.

 

If you feel its more work and that makes you feel like the bigger man than by all means go for it. The one thing I have learned over the last few years is there is no reasoning with you on this. In your mind you are the best and everyone else is just stupid. I'm sorry you need do that to validate your self worth but some day you will realize there is more than one way to do things.

 

By the way you don't need to be an @ss to every one on here. It doesn't make you look very professional.  

 

Ben, I'll try my best to give you the benefit of the doubt here.

 

Where did I mention that I have problems with every machine I encounter? Are you referring to my finding that machines are usually out of square?

 

Have you ever squared a machine before? Do you understand why squareness is critical in maintaining tolerances? If your first step isn't measuring square, then your cor numbers will be out along the stroke of the offending axis/axis'. I think this is the thing most cor guys don't grasp, and if you don't run the type of parts that are effected by it, it's easy to over look. I'll admit I overlooked it until it bit me in the @ss.

 

You could just answer some simple questions or add some factual information, data , examples etc, but instead you avoid it and just accuse me of ridiculous stuff. If you don't understand the topic at hand or don't have anything factual or useful to add, why not just keep it to yourself?

[BenK]

Ben, I'll try my best to give you the benefit of the doubt here.

 

Where did I mention that I have problems with every machine I encounter? Are you referring to my finding that machines are usually out of square?

 

Have you ever squared a machine before? Do you understand why squareness is critical in maintaining tolerances? If your first step isn't measuring square, then your cor numbers will be out along the stroke of the offending axis/axis'. I think this is the thing most cor guys don't grasp, and if you don't run the type of parts that are effected by it, it's easy to over look. I'll admit I overlooked it until it bit me in the @ss.

 

You could just answer some simple questions or add some factual information, data , examples etc, but instead you avoid it and just accuse me of ridiculous stuff. If you don't understand the topic at hand or don't have anything factual or useful to add, why not just keep it to yourself?

 

First off, you only asked me one question before this post which I responded to, so don't get your panties in a bunch. Second, you assume I don't understand what you are talking about based on your presumption that everyone besides you is stupid. The way you talk to people is ridiculous. You want to talk about what us COR guys don't grasp, how about what you don't grasp. We get it to work accurately. If you can't that's not our fault. I could add all the case studies and information in the world but you would still be pig headed and tell me I don't know what I'm doing. I'm fine if you don't believe me, I don't need you to approve of my method to validate my self worth. You have no clue the work I do so do yourself a favor and stop assuming. You want to attack people on here because we have a different way of doing things, why don't you look within and ask yourself why you feel the need to belittle others just because we don't bow down to you as the horizontal god you think you are. You have a method you are comfortable with, fine. My method works very well for me, I don't care if you approve.

[Sticky]

First off, you only asked me one question before this post which I responded to, so don't get your panties in a bunch. Second, you assume I don't understand what you are talking about based on your presumption that everyone besides you is stupid. The way you talk to people is ridiculous. You want to talk about what us COR guys don't grasp, how about what you don't grasp. We get it to work accurately. If you can't that's not our fault. I could add all the case studies and information in the world but you would still be pig headed and tell me I don't know what I'm doing. I'm fine if you don't believe me, I don't need you to approve of my method to validate my self worth. You have no clue the work I do so do yourself a favor and stop assuming. You want to attack people on here because we have a different way of doing things, why don't you look within and ask yourself why you feel the need to belittle others just because we don't bow down to you as the horizontal god you think you are. You have a method you are comfortable with, fine. My method works very well for me, I don't care if you approve.

 

 

I think it's time to just deal with the facts here instead of this made up non sense.

 

1. I asked if you could explain how you came to the conclusion I don't know how to find cor, you didn't. If you don't want to that's fine!

2. I'm not sure where you get the idea that I think I'm smarter then everyone else, and everyone else is stupid. Is it just because I called programming off cor only lazy? What is it that is getting you and Ron so worked up? Have you noticed that Mick and Newbee disagreed with Ron? Why aren't you making up ridiculous claims about them and claiming they are attacking people?

3. The whole thread is about using either multiple offsets or cor only, I've pointed several ACTUAL REAL DIFFERENCES. No one is stopping you from doing the same.

4. I can get cor to work, even somewhat accurately (it's a relative term), but it's not a robust and flexible process, as I've given a few examples of (and some of the other members have as well) of why it is less than optimal, and certainly not in "good" practice. All you have done to dispel this is claiming I can't find cor

5. Could you please find one "case studies and information in the world" about using cor only being better practice for programming a hmc instead of multiple work offsets?

6. I haven't attacked anyone here, but please feel free to provide an example, you make it sound like it's something I do all the time, so it should be easy for you to find an example, right?

7. I've said a few times that there are lot's of cases where cor only programming works, and if it works for you go for it. Several times in this thread even. So I don't really understand why you are so bent out of shape about this.

[Zoffen]

ya i love arguing on the internet! 

[BenK]

I think it's time to just deal with the facts here instead of this made up non sense.

 

1. I asked if you could explain how you came to the conclusion I don't know how to find cor, you didn't. If you don't want to that's fine!

2. I'm not sure where you get the idea that I think I'm smarter then everyone else, and everyone else is stupid. Is it just because I called programming off cor only lazy? What is it that is getting you and Ron so worked up? Have you noticed that Mick and Newbee disagreed with Ron? Why aren't you making up ridiculous claims about them and claiming they are attacking people?

3. The whole thread is about using either multiple offsets or cor only, I've pointed several ACTUAL REAL DIFFERENCES. No one is stopping you from doing the same.

4. I can get cor to work, even somewhat accurately (it's a relative term), but it's not a robust and flexible process, as I've given a few examples of (and some of the other members have as well) of why it is less than optimal, and certainly not in "good" practice. All you have done to dispel this is claiming I can't find cor

5. Could you please find one "case studies and information in the world" about using cor only being better practice for programming a hmc instead of multiple work offsets?

6. I haven't attacked anyone here, but please feel free to provide an example, you make it sound like it's something I do all the time, so it should be easy for you to find an example, right?

7. I've said a few times that there are lot's of cases where cor only programming works, and if it works for you go for it. Several times in this thread even. So I don't really understand why you are so bent out of shape about this.

 

Yes your right, all I have is made up nonsense. I guess that means the program I just finished is in no way accurate because you said it isn't possible. I responded to your question, if you have this kind of trouble with every machine you touch maybe you need to re-evaluate what you are doing. I would be glade to post specifics about work I have done and been successful with COR but I am bound by a NDA. Go ahead and assume what ever you wish about me. I know what I have done and will continue to do. There is no convincing you as stated earlier so why should I waste my time. This topic has been brought up many times and always ends up with you attacking people like this so I'm done. 

[Sticky]

Lol, this is like arguing with Micro.

[mike93]

using both at the same time is nice you can start with all offsets the same but if there is problems all faces can be adjusted independently to adjust for anything that may happen

[jlw™]

Fwiw, I use both depending on the situation. Nothing programming is black and white in machining and there is always more than to split the wig.

[BenK]

Lol, this is like arguing with Micro.

 

What is your problem? I have told you I do COR on a regular basis on close tolerance parts but that's not good enough for you. Face it there is nothing I could say to reason with you so no i'm not going to go through your silly list of questions. You have a problem with that i'm sorry but that's your problem. 

[Newbeeee™]

 If your first step isn't measuring square, then your cor numbers will be out along the stroke of the offending axis/axis'. I think this is the thing most cor guys don't grasp, and if you don't run the type of parts that are effected by it, it's easy to over look. I'll admit I overlooked it until it bit me in the @ss.

 

Geometry is what bit us at the old place (the 4x Horis).

The factory floor was sh!te and some machines were geometrically squarer than others. Machines were placed across different slabs rather than on one slab - and add to this the water table...

Throw into the mix that all weren't perfect at rotating around the centre line, it was a nightmare.

So this is why we went this way and it worked well.

 

For our rotary work (small bar in a 5c collet) we'll go COR and depending upon tolerances, go one datum or multiple for each index.

For bigger work in a vice on the 4th, we'll go a separate G5# offset for each of the 6 faces (talking prismatic 6 face 'boxes' here).

We'll use top (outside) face at each index for Z0, so all Z moves are Z- (into the part). This works better for us for retracts and clearances etc, as a 1mm positive value will definitely be clear of the face, rather than a large value from the COR.

Yes I know we could use incremental for clearances more, but absolute values work best for us (but I am a Newb at this).

For indexing, we'll just add a ref point (Z) value on the operation before indexing to clear.

 

Group hug everyone?

[mayday]

Well, I guess I'm also a lazy programmer.

not my job to decide offsets, that's for the setup/operator

35 years of doing it all wrong on 4 and 5 axis machines. DOH !

[Newbeeee™]

Well, I guess I'm also a lazy programmer.

not my job to decide offsets, that's for the setup/operator

35 years of doing it all wrong on 4 and 5 axis machines. DOH !

You're never too old to learn Wes ROR!

[specv]

IMHO I let the print decide, if its an aerospace part with 30 datums or a simple part with 3 datums. It really depends on the part. Mill turn is COR for me and most 4th axis is the part datums. I usually pick only 1 offset up and let angle macros calculate the rest of the offsets. If the part is on center of the table I use COR.

 

I have found over the years that this topic it really a hotly debated one with no one right in all instances. I remember my tech school instructor always said "give the same print to 10 machinists and you'll get 10 different processes" this is no different. Everyone here has different parts, machines and requirements and there is not one way that works best for all.

[Bob W.]

Geometry is what bit us at the old place (the 4x Horis).

The factory floor was sh!te and some machines were geometrically squarer than others. Machines were placed across different slabs rather than on one slab - and add to this the water table...

Throw into the mix that all weren't perfect at rotating around the centre line, it was a nightmare.

So this is why we went this way and it worked well.

 

For our rotary work (small bar in a 5c collet) we'll go COR and depending upon tolerances, go one datum or multiple for each index.

For bigger work in a vice on the 4th, we'll go a separate G5# offset for each of the 6 faces (talking prismatic 6 face 'boxes' here).

We'll use top (outside) face at each index for Z0, so all Z moves are Z- (into the part). This works better for us for retracts and clearances etc, as a 1mm positive value will definitely be clear of the face, rather than a large value from the COR.

Yes I know we could use incremental for clearances more, but absolute values work best for us (but I am a Newb at this).

For indexing, we'll just add a ref point (Z) value on the operation before indexing to clear.

 

Group hug everyone?

My floor is crap and that is a huge reason I went with Makinos for the horizontal machines.  They sit on three points so the floor can't incuce any twist in the machine.

[Sticky]

What is your problem? I have told you I do COR on a regular basis on close tolerance parts but that's not good enough for you. Face it there is nothing I could say to reason with you so no i'm not going to go through your silly list of questions. You have a problem with that i'm sorry but that's your problem. 

 

So you just want us to believe you because you say so?

 

There are actually plenty of things you can say to reason with me, like using factual information. Saying "because this is the way I always do it", while factually correct offers zero information as to the how's and why's, which is what this thread is all about. All you've done in this thread is whine and accuse, if you don't want to add any useful information just move on.

 

IMHO I let the print decide, if its an aerospace part with 30 datums or a simple part with 3 datums. It really depends on the part. Mill turn is COR for me and most 4th axis is the part datums. I usually pick only 1 offset up and let angle macros calculate the rest of the offsets. If the part is on center of the table I use COR.

 

I have found over the years that this topic it really a hotly debated one with no one right in all instances. I remember my tech school instructor always said "give the same print to 10 machinists and you'll get 10 different processes" this is no different. Everyone here has different parts, machines and requirements and there is not one way that works best for all.

 

it's definitely a lot easier to manage cor only when you just have one part of the table. Which is what I'll do for non critical stuff on the horizontal or mill turns. As soon as I have to deal with tool blending, in process inspection etc though I'll switch back to multiple offsets even if it is just one part.

[crazy^millman]

 

it's definitely a lot easier to manage cor only when you just have one part of the table. Which is what I'll do for non critical stuff on the horizontal or mill turns. As soon as I have to deal with tool blending, in process inspection etc though I'll switch back to multiple offsets even if it is just one part.

 

So let me get this straight when you want a simple easy way to program your machine for one part you do it exactly the way I said? The way I called a good practice is a way you program parts sometimes?

 

:wallbash: :wallbash: :wallbash:

 

And I am the one that needs to work on my grammar? I do have dyslexia, but I think I don't sound like an idiot and glad at least I have been able to help others for years using my bad grammar.

[Sticky]

So let me get this straight when you want a simple easy way to program your machine for one part you do it exactly the way I said? The way I called a good practice is a way you program parts sometimes?

 

:wallbash: :wallbash: :wallbash:

 

If it's loose and easy, then yes, sometimes I will, I don't do much of that type of work though so it is rare. I explained on the first page the application differences between the two methods

[BenK]

So you just want us to believe you because you say so?

 

There are actually plenty of things you can say to reason with me, like using factual information. Saying "because this is the way I always do it", while factually correct offers zero information as to the how's and why's, which is what this thread is all about. All you've done in this thread is whine and accuse, if you don't want to add any useful information just move on.

 

 

I would be glad to have this discussion with a rational and respectful person. Sadly you have proven to be neither. I guess it was my mistake being sucked into another thread about horizontal programming after seeing your name. You turn them all into an argument and I should have known better. Feel free to believe whatever you would like, I know what work I do and how well my methods work for me and that's good enough for me.

[Newbeeee™]

Ohhhhhhh, that's done it.

And I was looking forwards to a group hug. :sigh:

[Zoffen]

Man you guys all got it wrong. 

 

I use g53 machine co-ordinate to program everything in lol!!!! This is all ridiculous! Don't feed the trolls!

[Sticky]

Man you guys all got it wrong. 

 

I use g53 machine co-ordinate to program everything in lol!!!! This is all ridiculous! Don't feed the trolls!

 

No you have it all wrong, that will never work, you HAVE to use G92!!!111111