What's new
What's new

Esprit vs NX programming time with full sim

pharaun159

Plastic
Joined
Feb 2, 2016
I need some help here. Our company has been paying for 7 seeds of NX for around 5 years. Only one of which has been being used regularly by our 5 axis DMU programmer. Our other programmers refuse to use it and keep using mastercam to program simple mills and NLX lathes.

I was hired to program a 5 axis Swiss lathe using Esprit. The post was a mess but i worked with the author and have a reliable post now. The sim has worked since day one

For background Im certified in Gibbscam and esprit. I also have some background in creo, bobcam and mastercam. Im also a big fan of Mazateol for basic stuff.

I program lathes, mills, millturns, and swiss. Primarily swiss and mill turn. Usually 5axis.

Now that i have this stuff up and going, company wants to make some progress on getting the NX software working for their lathes, as they have invented alot into it. They want full, working simulation of the live machine environment. They purchased a kit from DMG mori for our NLX lathes.... Its a hot mess. I haven't seen it work yet, i haven't met anyone that can get it to work. We even brought a 30year veteran from Saratech in to assist. We made progress with him, but he was still struggling with it. It will post code but cant sim. One of the programmers managered to get the sim up and working up untill a part transfer but it breaks after that.

Even looking online i can't find anyone that actually uses nx to regularly full sim their lathe parts. They usually seem to build their geo and then sim the tool path, but they dont bother setting up all the kinematics for the sim. When i have gotten people to show me how to set it up... Its such a time consuming process, that i cant even begin to see how NX is competitive in anyway shape or form in the lathe market.

Now to be clear, im not the kind of guy that balks at anything new and drags his feet trying to stick to the old way. I see some serious benefits for milling operations in NX that remind me alot of things i liked in Gibbscam. Id like to see this get off the ground here.

As it stands now.. i have days into programming one part in NX that i can program, full sim, and post in quite literally do in 20min in esprit.

The company is willing to spare no expense for training to get me up to speed on this. I just need a direction. Where to evem start. I feel like it cant possibly be as hard as i feel like im making it out to be. Does NX sim require a bit more drafting and design expertise to handle managing these kinds of files?
 
I need some help here. Our company has been paying for 7 seeds of NX for around 5 years. Only one of which has been being used regularly by our 5 axis DMU programmer. Our other programmers refuse to use it and keep using mastercam to program simple mills and NLX lathes.

I was hired to program a 5 axis Swiss lathe using Esprit. The post was a mess but i worked with the author and have a reliable post now. The sim has worked since day one

For background Im certified in Gibbscam and esprit. I also have some background in creo, bobcam and mastercam. Im also a big fan of Mazateol for basic stuff.

I program lathes, mills, millturns, and swiss. Primarily swiss and mill turn. Usually 5axis.

Now that i have this stuff up and going, company wants to make some progress on getting the NX software working for their lathes, as they have invented alot into it. They want full, working simulation of the live machine environment. They purchased a kit from DMG mori for our NLX lathes.... Its a hot mess. I haven't seen it work yet, i haven't met anyone that can get it to work. We even brought a 30year veteran from Saratech in to assist. We made progress with him, but he was still struggling with it. It will post code but cant sim. One of the programmers managered to get the sim up and working up untill a part transfer but it breaks after that.

Even looking online i can't find anyone that actually uses nx to regularly full sim their lathe parts. They usually seem to build their geo and then sim the tool path, but they dont bother setting up all the kinematics for the sim. When i have gotten people to show me how to set it up... Its such a time consuming process, that i cant even begin to see how NX is competitive in anyway shape or form in the lathe market.

Now to be clear, im not the kind of guy that balks at anything new and drags his feet trying to stick to the old way. I see some serious benefits for milling operations in NX that remind me alot of things i liked in Gibbscam. Id like to see this get off the ground here.

As it stands now.. i have days into programming one part in NX that i can program, full sim, and post in quite literally do in 20min in esprit.

The company is willing to spare no expense for training to get me up to speed on this. I just need a direction. Where to evem start. I feel like it cant possibly be as hard as i feel like im making it out to be. Does NX sim require a bit more drafting and design expertise to handle managing these kinds of files?
saratech is ok at CAD stuff with NX, but i've found them to be lacking in the CAM side of things. if they're willing to spend money to get it set up right, go with Design Fusion, Jean there is a fucking ninja with NX posts/simulation. i personally know people that are doing 5 axis mill/turn with NX, using posts/sim from them. i have 2 posts/sim kits from them as well, but i dont do any turning, only 5 axis.

while initial setup is complicated - once its set up, its super easy, especially if you spend a bit of time to build some templates.
 
I use NX daily to program lathes, mills and millturns. With sims and it still impresses me with what can be done, even after 15 years of use.

Have a look at these guys:

Ralius Industries
Modulus Cam

Also, put out a request on the Siemens NX CAM forum. Plenty of expert help there.

Also there is this guy
NCMatic

Your request should be a cakewalk.

Also look at these guys youtube channel
CAM Division
 
We been looking at having Swoosh build us a kit. Ill check out the one you mentioned
I use NX daily to program lathes, mills and millturns. With sims and it still impresses me with what can be done, even after 15 years of use.

Have a look at these guys:

Ralius Industries
Modulus Cam

Also, put out a request on the Siemens NX CAM forum. Plenty of expert help there.

Also there is this guy
NCMatic

Your request should be a cakewalk.

Also look at these guys youtube channel
CAM Division
Really appreciate it!
 
The Key to making NX CAM work on complex machine tools is that it is absolutely critical you have a CAM guru. Most of the names mentioned here quality (Ralius, NCMatic, Swoosh). Even within those groups, different folks specialize in different things.

I think folks underestimate just how complicated a piece of software a simulation kit for a machine tool is. CAMPlete took on the task, and even they only have a fraction of the available machine-tools covered; each kit takes months of work, even with direct access to the machine tool builder. There are so many corner cases that need to be gone through, control settings that must be managed, canned cycles, motion events, etc etc. My feeling on a really good mill/turn sim kit is kind of like whenever I see a telephone cable trunk opened up; it is one of nature's miracles these things ever fucking work.

Siemens is getting into the game directly, but this will take years. With Post Configurator, they now formed a post/simulation kit team that is slowly going through and banging out factory simulation kits.

In the meantime, call one of the folks above and prepare to spend.
 
Also look at the sample machine kits shipped with NX.

The simulations cover most kinematic arrangements except for Y axis turret lathes which will most likely change, plus, post hub machine kits definitely exist to cover this gap.

Most have Siemens, Heidenhain, Fanuc and Okuma posts that are good starting points and can be tweaked by your reseller to suit your options or M codes. Or, you can reuse the kits and build your own post from the ground up. We have a cantankerous Okuma that I built a post for and once you know how it really opens possibilities.

This is an old video but it does show how deep you can dive;

NX Sim 08

and if you have access to NX look at the sample setup part for Sim 15

And for good measure synch manager.

Triple turret lathe
 
I use NX daily to program lathes, mills and millturns. With sims and it still impresses me with what can be done, even after 15 years of use.

Have a look at these guys:

Ralius Industries
Modulus Cam

Also, put out a request on the Siemens NX CAM forum. Plenty of expert help there.

Also there is this guy
NCMatic

Your request should be a cakewalk.

Also look at these guys youtube channel
CAM Division


Update on my quest here.

We have a DMG Mori purchased machine Kit. We are still struggling to get it to work quickly. It seems, that as far as things stand... NX is just crazy slow when it comes to lathe work. Putting tool blocks on the turret, setting the snap points, assigning where those tools sit on the blocks. We have over 230 different tool block combinations just for our normal production lathe work. Its a full time job just getting all this setup just to sim correctly. The Mori kit has about 60 expressions built in to help to "quickly" get a part loaded into the the chuck and make a copy and position it for the sub side. I just had a meeting with Swoosh and DMG trying to get his going. Every question went along the lines of "how to do this".. "oh thats easy, you just do... (insert 30 min explanation)".

When I first came to this company, a previous programmer left them in a bad spot. They had 21 SKU's (630) that needed programmed an ran in 6 months. I had no experience with the particular machine they needed to be run on. They bought Esprit specifically for this machine. I had self taught myself Esprit using the Cam Wizard training videos in one month (been using it for about 2 years now). We pulled it off. Customer was happy and launch was a success.

I can not see that scenario working with NX. Things that take a few minutes to do in Esprit, can literally take a day in NX. I agree its 5 axis tool paths and modeling is better then Esprit, but that only encompasses less then 1% of my work load. Feels more like my job has gotten harder in every capacity then it was before and we have gained nothing for it. I've talked to guys with 20-30 years NX experience, and they all admit that Esprit kicks NX ass on program turn around on Mill/turns, custom tool paths, and tool management. These are huge parts of programming multichannel lathes, sliding head stock lathes, and mill/turns. We are a job shop. Time is money.


Ok.... Bitch fest over.


The problem is, this software is not intuitive to me at all. Its such a "wide open" space that its a journey to get anywhere. Im sure once i tailor it to what i need, and build my UI the way i want it, that will help. However, I dont even know where to start. I feel like I have to flip a hundred switches to do what used to be one switch flip. We have another programmer on the mills trying to learn NX and he is struggling as well. So, seeing as this is what the company wants... I have to make this work somehow.

What is the best training out there? Anything like CAMWizard? Company will send us both.
 
some of the channels i follow for nx content

wish i could help you, but i've only done mill work with NX so far.
if you have whatsapp, i can invite you to a small NX group we have, couple guys here do mill/turn and might be able to help.
 
Every question went along the lines of "how to do this".. "oh thats easy, you just do... (insert 30 min explanation)".
Oh man this is one of my biggest gripes with NX. Just looking up how to do things in the Post Builder returns things like,

"Well all you need to do is (explanation that requires a masters in computer science) then you ( requires knowledge built up from 20 years using .TCL coding) then you make a custom journal to change all of these values."

How do journals work?

"Oh that's simple, (insert 30 page dissertation from someone who has been programming since UG1)"
 
Oh man this is one of my biggest gripes with NX. Just looking up how to do things in the Post Builder returns things like,

"Well all you need to do is (explanation that requires a masters in computer science) then you ( requires knowledge built up from 20 years using .TCL coding) then you make a custom journal to change all of these values."

How do journals work?

"Oh that's simple, (insert 30 page dissertation from someone who has been programming since UG1)"
Dude, Its killing me.... I was super excited to learn this...but honestly.. If i cant make some progress with this soon...and shit hits the fan with a customer wanting stuff "now".. Its gonna go the way of.. buy a post for esprit or find another programmer to do this. This software does not fit how i think about things at all.

I think it has a lot to do the methodology of all the "constraints".Snap this to this, on top of this, then connect this to this, then center this to this..its....messy. There is so much shit on my fing screen. A wall of shit for the assembly. A wall of shit for the tool pathing. A wall of shit for the part. Half my time is just trying to trying to get shit right for the SIM, then the rest is actual programming..and heaven forbid someone opens and messes with the model im working on. When i buy a kit from esprit, the machine enviornment is all built and hidden. Its not part of my "file". Its componants are not visible in my ui. Its not needed information. If fro some reason there is an issue with the machine enviornment, i can open the models and fix it... or just go.. hey esprit...fix this. and poof. Update.

In esprit i load the part in, click one face and click orient to x axis..Now i'm up and programming.In regards to handling lathe tool blocks. Triple holders and double stick holders.. All that stuff is easy. I have a chart with all the center positions on all the blocks. I literally just type in the coordinates of those positions when i put a tool in. In NX, i first gotta open the model of the tool. Then i gotta set up its orientation so when it loads into the assembly the right way. Trial and error that a few times.. ok now load in the tool block and add the tool.. shit the tool is sticking out 4in from the face of the tool block, because some shmuck modeled the snap point of the block from the face. Go back into my stick tool model. Move the snap point to the opposite end of the tool and set up a distance constraint so i dont have to fuck with it again. then try again.

In esprit.. shit some smuck modeled this block wrong.. oh well... X - 4in. Poof. Tools in place.


A lot of nx guys have shown me stuff they do to be quick.. most of which seems to be ignoring the sim entirely and just programming and posting the code. At that point, why bother paying the money for NX. You coud do that with fusion 360 or mastercam.
 
Dude, Its killing me.... I was super excited to learn this...but honestly.. If i cant make some progress with this soon...and shit hits the fan with a customer wanting stuff "now".. Its gonna go the way of.. buy a post for esprit or find another programmer to do this. This software does not fit how i think about things at all.

I think it has a lot to do the methodology of all the "constraints".Snap this to this, on top of this, then connect this to this, then center this to this..its....messy. There is so much shit on my fing screen. A wall of shit for the assembly. A wall of shit for the tool pathing. A wall of shit for the part. Half my time is just trying to trying to get shit right for the SIM, then the rest is actual programming..and heaven forbid someone opens and messes with the model im working on. When i buy a kit from esprit, the machine enviornment is all built and hidden. Its not part of my "file". Its componants are not visible in my ui. Its not needed information. If fro some reason there is an issue with the machine enviornment, i can open the models and fix it... or just go.. hey esprit...fix this. and poof. Update.

In esprit i load the part in, click one face and click orient to x axis..Now i'm up and programming.In regards to handling lathe tool blocks. Triple holders and double stick holders.. All that stuff is easy. I have a chart with all the center positions on all the blocks. I literally just type in the coordinates of those positions when i put a tool in. In NX, i first gotta open the model of the tool. Then i gotta set up its orientation so when it loads into the assembly the right way. Trial and error that a few times.. ok now load in the tool block and add the tool.. shit the tool is sticking out 4in from the face of the tool block, because some shmuck modeled the snap point of the block from the face. Go back into my stick tool model. Move the snap point to the opposite end of the tool and set up a distance constraint so i dont have to fuck with it again. then try again.

In esprit.. shit some smuck modeled this block wrong.. oh well... X - 4in. Poof. Tools in place.


A lot of nx guys have shown me stuff they do to be quick.. most of which seems to be ignoring the sim entirely and just programming and posting the code. At that point, why bother paying the money for NX. You coud do that with fusion 360 or mastercam.
how do you usually position parts in an assembly?

building tool and machine libraries is tricky, no doubt. however, once you do set them up - i've found programming different parts to be a breeze, at least 5 axis mill stuff.
 
how do you usually position parts in an assembly?

building tool and machine libraries is tricky, no doubt. however, once you do set them up - i've found programming different parts to be a breeze, at least 5 axis mill stuff.
It seems to entirely depend on how "who ever came before me" loaded different blocks into the data base. So i kinda have that working against me. Because they are not all done the same.

The short answer is..I dont know. Im really not sure what i'm even looking at.
Mill turns with turrets are not like a 5 axis mill. With a 5 axis mill, when you load your part into your template, your expressions can be setup to just snap that part to your fixturing. In regards to tools they all snap to the same location because its a mill. With a turret lathe you generally have 12 snap locations for tools. Ive had specialty turrets that had 36 locations. 12 outter locations, 12 "half" positions in between, with another 12 "face postions" for ID tools. So thats 36 snap locations...just for the the pod locations. And 24 of those positions can have things oriented towards the main spindle or towards the sub spindle, while the other twelve can only face the main. If you put a double stick holder in there, you have to snap the block to the turret position, and then create 2 snap points on the block to snap to. Then that same block can be oriented to face the main spindle or the sub spindle. And the stick tools themselves can be oriented in 2 postition per spot.. per orientation. So 8 different orientations. just for that one block. I could potentiall build an assembly of a block with every "common" tool "in it" and set a distance constraint off every tool end to control the stickout lengths. Then i could just load a block in, and turn off what i dont need. But im not sure how that will fly if i load that some assembly in two or more tool locations.

I'm probably overcomplicating things, but i was tasked with making sure the lathes were simulating correctly because of how easy it is to crash them. The general consensus that i have been given so far, has been to ignore sim.


I feel like i have to learn the modeling and design side of NX before i can even start learning how to setup a tool library to use this.
 
how do you usually position parts in an assembly?

building tool and machine libraries is tricky, no doubt. however, once you do set them up - i've found programming different parts to be a breeze, at least 5 axis mill stuff.
We have a guy programming our DMU mills and hes having no issues. For my previously stated reasons. They just are not as complex. Put blank in vise, center, and pull tools from library... go.

Our NLX lathe are pushed to the max. We have 210 different combinations of tool blocks, and 4 different chucks.. on of which is a programmable Eccentric Chuck for turning off-center. Our lathes are more trunions then lathes lol.
 
It seems to entirely depend on how "who ever came before me" loaded different blocks into the data base. So i kinda have that working against me. Because they are not all done the same.

The short answer is..I dont know. Im really not sure what i'm even looking at.
Mill turns with turrets are not like a 5 axis mill. With a 5 axis mill, when you load your part into your template, your expressions can be setup to just snap that part to your fixturing. In regards to tools they all snap to the same location because its a mill. With a turret lathe you generally have 12 snap locations for tools. Ive had specialty turrets that had 36 locations. 12 outter locations, 12 "half" positions in between, with another 12 "face postions" for ID tools. So thats 36 snap locations...just for the the pod locations. And 24 of those positions can have things oriented towards the main spindle or towards the sub spindle, while the other twelve can only face the main. If you put a double stick holder in there, you have to snap the block to the turret position, and then create 2 snap points on the block to snap to. Then that same block can be oriented to face the main spindle or the sub spindle. And the stick tools themselves can be oriented in 2 postition per spot.. per orientation. So 8 different orientations. just for that one block. I could potentiall build an assembly of a block with every "common" tool "in it" and set a distance constraint off every tool end to control the stickout lengths. Then i could just load a block in, and turn off what i dont need. But im not sure how that will fly if i load that some assembly in two or more tool locations.

I'm probably overcomplicating things, but i was tasked with making sure the lathes were simulating correctly because of how easy it is to crash them. The general consensus that i have been given so far, has been to ignore sim.


I feel like i have to learn the modeling and design side of NX before i can even start learning how to setup a tool library to use this.
i'm not familiar with espirit, but how would you do what you just described there? wouldnt you still have to position those tools in the correct locations/orientations? seems to me like you would, just a different way of doing it right?

look into 'arrangements' in NX, essentially you can build different configurations of your tool models/orientations, and turn on the one you need with a click of a button. this is all gonna take time to set up initially, absolutely. but pays off huge dividends in the long run
 
No simulation is gonna be ready "now" especially if you are learning on the fly. Don't worry so much about the demands of management and just develop some results and show them that you can do it. Then, like empower mentioned, you will probably have developed a basic template that you can work from.
 
i'm not familiar with espirit, but how would you do what you just described there? wouldnt you still have to position those tools in the correct locations/orientations? seems to me like you would, just a different way of doing it right?

look into 'arrangements' in NX, essentially you can build different configurations of your tool models/orientations, and turn on the one you need with a click of a button. this is all gonna take time to set up initially, absolutely. but pays off huge dividends in the long run
I would usually just need one copy of each tool block. Id write down the coordinates of each tool location. With one copy of each id tailor my machine setup to have one of every block in every pod and just turn them on or off as needed. Every kit ive bought has had all the models already supplied (because i request it). Sim models are just basic stls, so the file size stays small even with all of those models. As for the tools themselves, esprit sim doesn't give a shit where i place them. So i can just tell it a coordinate to go to. To adjust stickout, the tool pages have an override. Orientation is just a drop down menu when i load the tool in. Esprit figures out how to insert the tool based on the setting you set when you inport it. Using the tool in a different orientation doesnt change other files that are using that same tool.

Ive yet to find anyone who can explain how to build custom turning inserts and implement them in NX. The last guy i talked to said it really wasn't a feasible thing in NX. 70% of the tools i use for micro boring tools or custom turning inserts (custom as in, not standard 80deg, 55 deg, 35 deg inserts). They said NX is working on something to make it possible but its in its early stages.
 
I would usually just need one copy of each tool block. Id write down the coordinates of each tool location. With one copy of each id tailor my machine setup to have one of every block in every pod and just turn them on or off as needed. Every kit ive bought has had all the models already supplied (because i request it). Sim models are just basic stls, so the file size stays small even with all of those models. As for the tools themselves, esprit sim doesn't give a shit where i place them. So i can just tell it a coordinate to go to. To adjust stickout, the tool pages have an override. Orientation is just a drop down menu when i load the tool in. Esprit figures out how to insert the tool based on the setting you set when you inport it. Using the tool in a different orientation doesnt change other files that are using that same tool.

Ive yet to find anyone who can explain how to build custom turning inserts and implement them in NX. The last guy i talked to said it really wasn't a feasible thing in NX. 70% of the tools i use for micro boring tools or custom turning inserts (custom as in, not standard 80deg, 55 deg, 35 deg inserts). They said NX is working on something to make it possible but its in its early stages.
as far as your tooling blocks etc, arrangements are what you'd want to do. absolutely doable, just takes time to set up initially. insert wise i have no idea because i havent done turning so i'll leave that to those better familiar with it.
 
Update on my quest here.

We have a DMG Mori purchased machine Kit. We are still struggling to get it to work quickly. It seems, that as far as things stand... NX is just crazy slow when it comes to lathe work. Putting tool blocks on the turret, setting the snap points, assigning where those tools sit on the blocks. We have over 230 different tool block combinations just for our normal production lathe work. Its a full time job just getting all this setup just to sim correctly. The Mori kit has about 60 expressions built in to help to "quickly" get a part loaded into the the chuck and make a copy and position it for the sub side. I just had a meeting with Swoosh and DMG trying to get his going. Every question went along the lines of "how to do this".. "oh thats easy, you just do... (insert 30 min explanation)".

When I first came to this company, a previous programmer left them in a bad spot. They had 21 SKU's (630) that needed programmed an ran in 6 months. I had no experience with the particular machine they needed to be run on. They bought Esprit specifically for this machine. I had self taught myself Esprit using the Cam Wizard training videos in one month (been using it for about 2 years now). We pulled it off. Customer was happy and launch was a success.

I can not see that scenario working with NX. Things that take a few minutes to do in Esprit, can literally take a day in NX. I agree its 5 axis tool paths and modeling is better then Esprit, but that only encompasses less then 1% of my work load. Feels more like my job has gotten harder in every capacity then it was before and we have gained nothing for it. I've talked to guys with 20-30 years NX experience, and they all admit that Esprit kicks NX ass on program turn around on Mill/turns, custom tool paths, and tool management. These are huge parts of programming multichannel lathes, sliding head stock lathes, and mill/turns. We are a job shop. Time is money.


Ok.... Bitch fest over.


The problem is, this software is not intuitive to me at all. Its such a "wide open" space that its a journey to get anywhere. Im sure once i tailor it to what i need, and build my UI the way i want it, that will help. However, I dont even know where to start. I feel like I have to flip a hundred switches to do what used to be one switch flip. We have another programmer on the mills trying to learn NX and he is struggling as well. So, seeing as this is what the company wants... I have to make this work somehow.

What is the best training out there? Anything like CAMWizard? Company will send us both.
NX is a wide open space, intentionally, it's what makes it so good. However, that is what also makes frustrating for new users who are used to a set workflow and want to rinse and repeat. You can do that with NX but you * might * miss a really useful function by not trying out other options. I will just as often use a cavity milling toopath for facing away stock as I will use a dedicated prismatic facing toolpath.

Without knowing too much about your situation I will say that NX is a toolkit for you to build on. It will be whatever you want it to be, hence it can seem like there is no button to push to get a result. Templates, build your own to the way you want it. Post processors, build your own or have them built to your EXACT requirements, postbuilder ( old school now ) is for that reason such a black box for some and a godsend for others. Machines simulations the same. Speeds, feeds, everything can be tailored and is usually in the default installation, a blank slate. Teach a man to fish... etc etc.

Y axis lathes with sub spindles are complicated creatures. Millturns even more so. However I can confidently say that NX will comfortably handle it all, efficiently. Where I think Esprit has triumphed is that they have tried to preset things as much as possible for time poor programmers who have a boss demanding something now. I work for a job shop too, time is money and NX can be your best friend.

Tool blocks once you get one right the rest will follow. Start with simplest one, build your understanding and go from there. My tools snap to their mount positions instantly, no mess no fuss. Yours should too, X and Y mounts can confuse because you are using the CAD coordinate system not the machine axis layout. But you will get a grip soon enough.


Also, ask for help, frequently, 30 minute answers are because you have lots of choices. DM if you want
 








 
Back
Top