cancel
Showing results for 
Search instead for 
Did you mean: 

Modifing the CLS output for angle heads

Valued Contributor
Valued Contributor

I need to modify the cls file GOTO output when using angle heads by the y/z offset amount as defined in the tool. I have edited the clsf.tcl file in the proc GOTO_output to shift the cl numbers and offset them by the tool axis times the mom_tool_z_offset.

 

like this:

 

global mom_head_name mom_tool_z_offset

if {[info exists mom_head_name]} {

# MOM_output_literal "-------------------------$mom_head_name"

if {$mom_head_name != "NONE"} {

set mom_mcs_goto(0) $mom_mcs_goto(0)+(\$mom_tool_axis(0)*$mom_tool_z_offset)

set mom_mcs_goto(1) $mom_mcs_goto(1)+(\$mom_tool_axis(1)*$mom_tool_z_offset)

set mom_mcs_goto(2) $mom_mcs_goto(2)+(\$mom_tool_axis(2)*$mom_tool_z_offset)

}

}

 

this works for the first move of an operation but after that it reverts to the cl file numbers.

 

example:

the mom_tool_z_offset is equal to 1.25

Tool axis 0,1,0

 

GOTO/1.4500,-3.4500,13.0000,0.0000000,-1.0000000,0.0000000 (cl was 1.45,-2.2,13.0 so the shift worked)

CYCLE/DRILL,RAPTO,0.2500,FEDTO,-0.2500,IPM,10.0000

PAINT/COLOR,31

GOTO/1.4500,-1.9500,10.9250   (this line matches the cl file so the shift failed)

GOTO/-1.4500,-1.9500,10.9750   (this line matches the cl file so the shift failed)

CYCLE/OFF

PAINT/COLOR,186

RAPID

GOTO/-1.4500,-2.2000,13.0000   (this line matches the cl file so the shift failed)

 

Does anybody have an idea how to fix this? I am using a 3rd party post and I want the cl file shifted to the spindle centerline when using a right angle head. By doing this it will simplify the posting of the cl file.

 

Thank You!

 

 

44 REPLIES

Re: Modifing the CLS output for angle heads

Phenom
Phenom

Looks to me like you did it where it should work. I wonder if a turbo mode is putting out the goto points without coming through the motion events. In that case - you have two options - either disable the turbo mode or editing stuff in the MOM_before_output routine (from what I recall from threads.)

 

I did some searches and found these threads about this:

 

http://community.plm.automation.siemens.com/t5/NX-CAM/Export-CLS-with-tool-vector-AND-contact-vector...

 

http://community.plm.automation.siemens.com/t5/NX-CAM/clsf-tcl-customization-help/m-p/120812/highlig...

NX10.03
Windows 7 Pro

Re: Modifing the CLS output for angle heads

Esteemed Contributor
Esteemed Contributor

Note the lines that you indicated were wrong MAY be being output by the drill cycle code, rather than linear/rapid move code.  So you may have to "fix" this in another area of the CLS .tcl file

Ken Akerboom Sr CAx Systems Engr, Moog, Inc.
Production: NX10.0.3.5 MP5 + patch/TC11.2
I'd rather be e-steemed than e-diseaseled


Re: Modifing the CLS output for angle heads

Valued Contributor
Valued Contributor

Ok,

 By commenting the "set mom_kin_clsf_generation TRUE" line in the clsf.tcl file it makes my output for this issue come out as I wanted. But it messes with a lot of the rest of my output, so much so that this will not be acceptable. I will have to try the befor output proc to see what I can do with it. I am hoping I do not have to set up a Linked Post situation for the cls file.

Thanks guys, I will update this when I find something that works.

Thanks

 

Re: Modifing the CLS output for angle heads

Phenom
Phenom

Hi Randy,

The company I work for is interested in expanding the use of Icam posts to NX in (at least) situations where Creo is also used to program a machine. Currently there are quite a few Icam posts for Creo in some facilities - and they are trying to use NX also. They say that a single Icam post can be set up to use the existing posts for Creo - which passes a lot of stuff in different format - and apt word "cmd"s for stuff that NX would have custom udes for. I have asked those that are looking at it if the cl generator posts will need a lot of adjustment (like you are.) I know that NX's cl looks a lot different that the one from Creo. They say that Icam ships (and maintains) a cl post for NX. Is that true? Even if true - it seems like the cl generator will have to be changed - maybe even as much as a postbuilder G code post. I would not want to start entering Apt major/minor keyword commands all over - I would prefer to keep the look and feel in NX the same whether Icam or postbuilder post.

I am thinking that there will have to be a routine set up for each machine that looks up apt representation that Creo puts out to match udes. Maybe it would be better to make routines that parse the out of the box cl from NX and convert it. I suspect that advantages in ease of postbuilding in Icam may be offset by trouble of trying to get stuff through the cl interface. I think the people in our company don't know what they are getting themselves into.

Dan

NX10.03
Windows 7 Pro

Re: Modifing the CLS output for angle heads

Valued Contributor
Valued Contributor

I will try to answer your questions to the best of my limited ability. ICAM has a NX to CL file generator that is keyed to the format they like to use. I have had very little luck with it because we have customized how we use NX over the years. So I have been customizing the NX supplied cl file generator to put out the format ICAM likes. You are faced with two places to modify things, one is the cl file generator, and two is the ICAM post. The rule of thumb I have been using is if I want the modification to exist in all post I change the CL file generator but if the change is in just a few post I change the post. So far most of the modifications have been rather simple since most of the custom TCL code I programmed into my Postbuilder post can be scavenged for the cl file generator. It looks like in your case you will need to modify the cl generator to match the existing Creo format. While that is a pain to do understand this is only one post that you are modifying unlike having to change every post your company has. Once you have done it it should always work for you, at least that is how it is working for us.

   We are just getting started with this but we are seeing much higher quality out of the ICAM post than we get from our Postbuilder post. The linearization and intermediate points from ICAM are making our machines run smoother than our Postbuilder post. That may just be my lack of skill as a post writer but it looks like we will be making a higher quality product so the effort seems worth it.

I hope that helps

Re: Modifing the CLS output for angle heads

Phenom
Phenom

Hi Randy,

Great info. I understand the need to have control of and customize the cl post - I suspected it is necessary contrary to what people tell me that have only listened to sales people about the ease of it.

By "Linearization" - do you mean that ICAM will change to lines based on a tolerance where the cl sends through longer linear segments or circles - specifically when a rotary is involved? In these situations - I admit that post building in PB is a challenge. Best case - the control can be set to polar mode to handle it (ex. Siemens Transmit.) I have also broken down linear segments in MOM_linear_move for this case considering a tolerance and where rotation center is. There is a kinematic switch to turn the circles into points - but sometimes they are still too long. Also - this does not break the long linear moves down at all.

I think that use of sync manager or ISV would be impossible with a cl post. We have more and more multi-channel stuff. I will be watching for your comments on this - as I will probably be involved to do the same stuff you are doing with the cl generator.

Regarding the need to offset the goto points - is there no way to incorporate this in ICAM? I would think that ICAM would be aware of the RA shape and orientation - and handle (if the post must) any offsets associated with it (or the control with head comp - traori/G43.4 etc.)

My observation with the cl out of NX is that I think it carries a requirement of the abs axes orientation matches the machine axes orientation. If this is the case - and a program is finished when you find out that the models have to be rotated - I am sure you have the same experience as I do with that (may not have a good effect on the motion.) I think it is possible to correct this in the cl post (if it is a problem - I observed it when I started out with NX - maybe didn't know what I was looking at.)

Thanks, Dan

NX10.03
Windows 7 Pro

Re: Modifing the CLS output for angle heads

Valued Contributor
Valued Contributor

Dan,

  If I were to walk out today and buy Siemens NX and ICAM it would probably require minimum customization. Plus many companies use "Insert" statements to pass information to the post. But we have 10+ years of legacy programs and I have a phobia about having the programmers use "Insert" since it is passed thru to the post without verification. So I created ways around it using UDE's and special operation names such as RPM_10000_CLW. Being able to pass this to ICAM is where most of my work has been.

  I use the term linearization primarily for multiaxis moves where the post adds blocks between the NX generated cl points to better control the path. Most of my machines are programmed in RTCP (tool tip) programming and Postbuilder cannot linearize if it does not have a tool length or a pivot length. I realize that the machine should be able to handle this but I have had poor quality results depending on the machine control. What I did in Postbuilder was to put in a 10.00 inch pivot which allowed Postbuilder to calculate the needed extra points. I then subtracted the (10 inch pivot times the IJK vector values) from the XYZ values to transfer back to the tool tip for G43.4 or Traori. This has worked well for us but the 10 inch pivot is a best guess and is hard coded in the post.

  In ICAM you have commands you can hard code that allow you to output this information using a tolerance at tool tip and also control the side angle tolerance. This can also be controlled from the tape so if the standard is not good enough you can tighten it. I have found this to be far superior to the way I did it in Postbuilder. Multiaxis cuts that were +/-.05 inches before I made the Postbuilder change dropped to +/-.002 after I made the change. Those same cuts are now better than +/-.0005 in ICAM without the programmer changing anything.

  As far as Circles are concerned when ICAM needs to convert them to linear moves I am seeing more points than I got in Postbuilder and it appears the internal tolerance is better but I am not sure as to how to control that tolerance yet. I am sure there is a command somewhere in my 500 page manual that adjust this but I havn't looked for it yet.

  On your multi-channel  machines I was just looking at the changes for the newest version of ICAM and they have improved the "Merge" function so you may have much less trouble than you think here. We specialize in 5-axis milling and at this time we have no multi-channel machines in this facility so my experience here is none, sorry.

   On offsetting the CL points, It can be done in ICAM as long as you turn off RTCP (g43.3/Traori). You can then INSERT the TRAORI into the tape (bypassing ICAM) and get the output I am looking for but I was hoping to do this without needing to bypass the post. In RTCP I should not have to do this but again I have years of legacy programs that use this offset and no one wants to change all of them to put the offset into the control. We probably took a wrong turn years ago and we are kind of stuck on that road at this time.

Thanks

Randy

Re: Modifing the CLS output for angle heads

Phenom
Phenom

Randy,

Thanks again for all the insight on how ICAM works out of NX. I would have guessed that INSERTs for custom words to pass info through the cl would be the usual method - that way no changes need be made from machine to machine (in the cl generator) - just remembered by the programmer (differences kept in your head.) There would probably be issues with machine switching as well (a lot of apt statements in inserts to fix.) I would also think that there would be timing issues sometimes where the insert puts out an aptword in the cl where you don't want it yet - and then you have to build the correction of timing into the post if possible.

I understand the need to linearize to get better results and why you need to adjust the cl goto's. I would say that the nxpost design will allow all of this - but it is not out of the box - and I am sure a lot more work than ICAM. I made cl posts from scratch in VB for Catia output. At the time other post systems weren't offering rewind/prewind options - but they all do now. We never linearized with those (though) and depended on settings in cam for this type of thing (and the machine performance.) I wonder if tuning parameters at the machine could help with the need for so many points? If cam thinks a cutter is on the part to a tolerance - the rest is left for the machine to achieve it in those 5 axis moves with pivot lengths. Maybe some tuning or mode changes could help improve the tooltip path.

Good luck with the cl post modification.

Dan

 

NX10.03
Windows 7 Pro

Re: Modifing the CLS output for angle heads

Siemens Genius Siemens Genius
Siemens Genius

FYI, NX/Post should have used tool's flute length (if available) to perform linearization, even when you are posting for a dual-head machine and tracking at the tool tip. And, with Post Builder, you also have the choice of linearizing moves by interpolating (distributing) angles or tool-axis vectors.

 

Learn online





Solution Information