[jrmach]

Quote from mmoe

" everything else is just easier than dealing with inches, even compared to decimal inches."




Phooey


continue on gentleman

[SBC Cycle]

I drew up a 20mm thick part and set the Work Offset for G54 to X17.6, Y43.5, Z-68.5432 (to account for the stock). The workpiece is perfect and seems to be just where you would expect it (10mm off the X edge, flat against the table). The Z is still disconnected from the holder, arbor, and tool. In this screenshot, the total tool length including holder is 100mm. You can see the axis values and hopefully make some sense of them.

The gap between the holder (in blue) and the collet on the machine (light grey) is 50mm. Doesn't matter which holder I pick or if I remove the tool holder (BobCAD adds one by default even if you remove it).

EDIT: I added a second picture to give you an idea of how things can be moved around. I added a +50mm shift in the Z axis for the "holder_transform" and it moved the "Tool Set" up to the nose of your collet.

[mmoe]

I drew up a 20mm thick part and set the Work Offset for G54 to X17.6, Y43.5, Z-68.5432 (to account for the stock). The workpiece is perfect and seems to be just where you would expect it (10mm off the X edge, flat against the table). The Z is still disconnected from the holder, arbor, and tool. In this screenshot, the total tool length including holder is 100mm. You can see the axis values and hopefully make some sense of them.

The gap between the holder (in blue) and the collet on the machine (light grey) is 50mm. Doesn't matter which holder I pick or if I remove the tool holder (BobCAD adds one by default even if you remove it).

It's hard to tell what's going on there. The Z value is such that it would add to the work offset to produce a move to -80 in machine coordinates. I'll have to think on that a bit.......

[BurrMan]

So mmoe, you've got 8 things going on much too fast for me. I'm still stuck back at number 1, which was getting the sim to use work coords, or machine coords, per the multi axis posting page. I'm speaking with BC about this now.

You are doing a nice job, and I appreciate getting to play along because it helps the whole process for me.

I made a video of your latest machine setup (sbc beat me with a picture of it, and some description).

machine tool and limits - YouTube

In the video, I bring up the STL placement, and how to set your machine to work correctly, and also point out factors which are "tripping the limits" when we don't think they should. If our values don't add up, the limits are tripped. I also show how the tool setting is affecting it. But I didn't go into code output. It's just too daunting for me at the moment, because I have other questions I need answered before I can make sense of the numbers properly. With regard to the movelist and actual gcode created at the post. Maybe with this video and your discussion with sbc, you'll come out with some more understanding. "I" would define anything that stays on my machine, as an STL. With that though, I am also hoping I can put a transform on other components separately, so I can move them around, as you describe when you do setups. Like the 2 z heads, or shoving tool holders up into the spindle or what...

I havnt got that far yet. After I find out about the real machine coords question, I'll be able to get into moving parts with the transforms, and looking at the post output from that..

[SBC Cycle]

In the video, I bring up the STL placement, and how to set your machine to work correctly, and also point out factors which are "tripping the limits" when we don't think they should. If our values don't add up, the limits are tripped. I also show how the tool setting is affecting it. But I didn't go into code output. It's just too daunting for me at the moment, because I have other questions I need answered before I can make sense of the numbers properly. With regard to the movelist and actual gcode created at the post. Maybe with this video and your discussion with sbc, you'll come out with some more understanding. "I" would define anything that stays on my machine, as an STL. With that though, I am also hoping I can put a transform on other components separately, so I can move them around, as you describe when you do setups. Like the 2 z heads, or shoving tool holders up into the spindle or what...

I havnt got that far yet. After I find out about the real machine coords question, I'll be able to get into moving parts with the transforms, and looking at the post output from that..

I'm going to watch your video in a bit but initial thoughts are, I would also define anything that stays on (and moves with) my machine as an STL. The question becomes "where is Zero?". When I made my decision where Z zero would be, I took into account the way I set Work Offsets and how tool heights are stored. That may be why it appears wrong to me. In mmoe's post #33 he has a nice closeup sketch of his spindle. In that drawing, the "End of Spindle Collet" would be the face of the CAT40 taper spindle. The rest would all change dynamically by the holder and tool definition in BobCAD.

[BurrMan]

I'm going to watch your video in a bit but initial thoughts are, I would also define anything that stays on (and moves with) my machine as an STL. The question becomes "where is Zero?". When I made my decision where Z zero would be, I took into account the way I set Work Offsets and how tool heights are stored. That may be why it appears wrong to me. In mmoe's post #33 he has a nice closeup sketch of his spindle. In that drawing, the "End of Spindle Collet" would be the face of the CAT40 taper spindle. The rest would all change dynamically by the holder and tool definition in BobCAD.

Yes to all of that, but he also talks a lot about moving that collet in and out and such. This is where I can confirm, and I think a lot of the confusion is. "WE" are setting everything at 0 and don't see any issues. "He" needs to use Real machine coords" (I think.)

When you look at it, can you compare the code output of his normal machine and then the output with the holder transform shift applied, then also with the tool length + holder shift and where it actually cuts vs the movelist? I think you guys can talk turkey if you know those answers.

Just so you know. I really want to define transforms for both heads, after the initial machine is finished. Don't know if this is going to be possible or not yet.

[mmoe]

Burrman or SBC,
Would either of you be willing to try a toolpath with "Override Offsets" checked where you select the tool? I always check this box since I compensate for the tool in the work coordinates not the tool table. I wonder if that is what is causing the tool to be down so low (though I'd think it would be the other way around). The thing is, when I do it this way and simulate it without the machine, the tool is shifted exactly the way it would need to be, so if the machine was also shifted to the work coordinates as it should be (and would be in real life), then the tool would be inside the tool holder. Here's a quick shot of a simulation without the machine. You can see the origin is off to the side and higher than the tool. The position of the tool tip is precisely the amount of the offset, so with "Override Offsets" on, I think the tool tip is what is placed at 0,0,0 (or the work coordinates 0,0,0 where different from machine coordinates). You can see that the tool tip is at Z=10, which is what I set my Clearance Height to, so it appears to be correct.



And here you can see that there is a tiny dot of gold in the simulation of the machine. When zoomed in, that dot is an endmill that has it's tip exactly on the machine origin. It seems like changing the size of that endmill would only make it grow up, since it would appear to be justified to the Z0 plane. If it were then offset with work coordinates like the image above, it would be exactly in the correct place. Unfortunately I'm at a disadvantage for testing these things until January.

[mmoe]

I think I'll rework the STLs to align with the end of spindle, then add a tool holder (figured out how to define them last night) with the same geometry as the one currently in the machine model. I'm thinking that I can set the holder and arbor to be transparent in the macine definition, so the holder in the model may then be what's visible. That may set the tool to the correct location.

I'm on field trip duty with my 10yo today, but should have that available by 4pm pst.

[SBC Cycle]

Yes to all of that, but he also talks a lot about moving that collet in and out and such. This is where I can confirm, and I think a lot of the confusion is. "WE" are setting everything at 0 and don't see any issues. "He" needs to use Real machine coords" (I think.)

When you look at it, can you compare the code output of his normal machine and then the output with the holder transform shift applied, then also with the tool length + holder shift and where it actually cuts vs the movelist? I think you guys can talk turkey if you know those answers.

Just so you know. I really want to define transforms for both heads, after the initial machine is finished. Don't know if this is going to be possible or not yet.

I see what you are saying. I keep looking at the video of the machine and the way the head operates and I'm not grasping it yet.

LOL, you're going to get both heads cutting and I was proud of myself that I could make the pendant swivel to and fro I tried at some point to add another tool set and workpiece set to the Machine Definition but it would only use one of them. I didn't fool with it for too long. I'll try to help best I can. This is a fascinating machine to work with and is helping me get a better understanding of how it all fits together.

Have you figured a way to control an additional axis? Like I said about the pendant, I set it up as a rotation axis and could swivel it around by playing with the axis sliders but I would really love to know if there is a way to hack the Sim to throw in additional movement. I really have a serious use for it. Our new machines have a side mount tool changer arm and I've had a few close calls during tool changes. The arm by itself moves down a good 4.5" inches when changing tools, plus the length of the tool...I've had more than a few puckered bungholes where it was very close to a collision.

[BurrMan]

This is a fascinating machine to work with and is helping me get a better understanding of how it all fits together..

You are correct. It is covering many of the complexities.

LOL, you're going to get both heads cutting and I was proud of myself that I could make the pendant swivel to and fro I tried at some point to add another tool set and workpiece set to the Machine Definition but it would only use one of them. I didn't fool with it for too long. I'll try to help best I can.

Have you figured a way to control an additional axis? Like I said about the pendant, I set it up as a rotation axis and could swivel it around by playing with the axis sliders but I would really love to know if there is a way to hack the Sim to throw in additional movement. I really have a serious use for it. Our new machines have a side mount tool changer arm and I've had a few close calls during tool changes. The arm by itself moves down a good 4.5" inches when changing tools, plus the length of the tool...I've had more than a few puckered bungholes where it was very close to a collision

Not yet. I just got past step 1 I was stuck on. It's slow for me and will take some time to figure out what can/cant, should/shouldn't and can/should vs. DOESNT.

I need help to recognize the code output that SHOULD be happening, or I have to read each number and use a calculator! lol

[BurrMan]

So my problem on my step 1 issue is solved and was self inflicted. I should be able to get past the "why do the limits trip in this setup" question now, but it will take me a couple to test and sample.

My issue was I was so entrenched in working the definition and transforms, that I was changing the multi axis posting setting for real machine zero and work offset zero on THAT PAGE, which effectively doesn't change it for the current operation. It needed to be changed on the "PART" current settings.

Now that my code is output either by real machine zero or work offset, I can look at the next. I think mmoe has a list of about 8 or 9 by now!!!

Much thanks to the bc staff for helping the numbskull wasting their time. (the ole "why didn't I think of that" trick)

[mmoe]

Here's the latest machine definition update, which has shifted all the STLs down so that the end of the spindle is at the origin. I also created a tool holder, but I can't see a way to isolate it to share as a file. I could provide the whole tool holder file, but it would replace any existing toolholders that may have been created, so not really an option. Here's the specifications that I set the tool holder to, which should closely match the machine model's STL tool holder:

[BurrMan]

Here's the latest machine definition update, which has shifted all the STLs down so that the end of the spindle is at the origin.

I haven't looked at your latest setup yet. I need a break. I DO believe that you moving everything to the zero point will make it easier to get started and you can look at moving things back AFTER your up and running.

I'll jump back in in a bit, after I review some other things I'm looking at. NICE BUILD!

[BurrMan]

Have you figured a way to control an additional axis? Like I said about the pendant, I set it up as a rotation axis and could swivel it around by playing with the axis sliders but I would really love to know if there is a way to hack the Sim to throw in additional movement. .

This is what I'm looking at right now. It appears that they are using a 4X4 transformation matrix, so I don't think so with what we have. They have a manual_mode and autodetect_semiautomatic flags, but I don't think we have access to those (am looking).. Probably internal. But it could just be error on my part in my initial process.

I'll post back if I figure it out.

At the least, at this point, we he gets up and running, we can just make "2 machine files" with the Z heads configured to represent the switch in the sim. Then he can just select which machine to sim with to show which version. It sure would be nice to make that selection with a work coord or something though....

Anyway.... Fun.

[SBC Cycle]

I got tied up as well. I'm going to be working on my machine definition today as well along with a little modeling. I still need to draw up a machine base yet which will give me a reference point to better understand what's going on.

[jrmach]

Don't they have somewhere to buy machines and accessories on the net somewhere.You know,.STL'S

[SBC Cycle]

For a Haas owner there are some resources. I've acquired a few models from Grabcad that were ok but with SolidWorks it's not difficult to sketch up a model, then create something they calll a "configuration". Once I have the VF2 modeled, it takes almost no time to create the VF3. Same with the Fadals, I have 4 different size machines (3016, 4020 extended Z, 4020 Standard, and 5020 extended Z).

I've noticed people keep close tabs on their modeling work and aren't so quick to hand them out. The Factory certainly won't offer them. I plan to release my models when I'm finished, no need to be stingy.

[mmoe]

For a Haas owner there are some resources. I've acquired a few models from Grabcad that were ok but with SolidWorks it's not difficult to sketch up a model, then create something they calll a "configuration". Once I have the VF2 modeled, it takes almost no time to create the VF3. Same with the Fadals, I have 4 different size machines (3016, 4020 extended Z, 4020 Standard, and 5020 extended Z).

I've noticed people keep close tabs on their modeling work and aren't so quick to hand them out. The Factory certainly won't offer them. I plan to release my models when I'm finished, no need to be stingy.

It would be cool to make a thread that has only links to machine definitions and their needed STLs for people to download. While I doubt there are 3 other people with my exact machine in this country, there may be plenty of people who would just enjoy doing virtual machining on machines they don't actually have. All the fun with none of the noise, dust/debris, or expense. Just nothing physical to show for the efforts. There are plenty of people who model things just for fun and they never intend them to be anything but digital ideas, so this would just take that one more step.

[SBC Cycle]

I agree. I also have a Machine Definition thread a few pages back where I shared my HRT-160 rotary files and XML for the definition. I asked if we could swap definitions just by dropping a new folder in the MachSim folder. I didn't get a reply but with your project I think we have discovered an answer, "yes!". If someone wants to make a new thread I will contribute what I've drawn and found on the net.

[mmoe]

Bringing this thread back from the dead. I've finally got updated to V26 with Simulation Pro, so I'm now able to test how my simulation actually works other than moving the axis around by hand.

First, I find that it worked out very well to have the tool holder modeled in the simulation. The reason for this is that when I created my tool holder in Bobcad using the "Milling Tool Holder Library", I could verify that the tool holder from the library is located in the exact same position as the STL I created in ViaCAD when I modeled the entire machine. From that point, the tool itself extends based on the tool parameters pretty much exactly as you would expect. At the end of the video below, you can see that the light grey STL from my model is intermixed with the dark grey tool holder generated by the Bobcad library file, indicating that they are for all practical purposes identical. IMHO, this could not have worked out better.

The tricky part was understanding how the work offsets relate to the simulation. My real machine has the work offsets configured so that the zero position is the top of the part, so obviously this zero position changes based on the job, but it's easy to set up. The offset is the difference from the machine zero to the work coordinate zero, so the values entered are the values needed to shift the tool that amount in each axis. In the Bobcad simulation, this does not happen quite the same way.

First, assume that you are using the top of the part as your Z zero. If you do something different than that, you'll have to account for the difference in relationship produced by whatever zero you choose. Let's also say that you don't have any offset entered in the machine setup. If that is the case, the machine coordinates will be used for X and Y as normal, but then the head will be raised to shift the top of the material to the machine Z0 position. This moves the head up by roughly the combined values of the tool protrusion length, the tool holder overall length and any Z value of the clearance plane, since the tool starts the simulation at the clearance plane. At first, this behavior seems confusing, but it will make more sense later.

So how do you set up the material/part to be where you want it? What I found is that I had to know the difference between the top of my table with no stock thickness (since this has to be subtracted based on each specific stock thickness) and the machine zero (which is the bottom of the collet nut on the spindle, and that's how I highly recommend you set it up). In my case, that value is 198.5mm as it sits in my current simulation, though I may make my wasteboard 1.5mm thinner just to make it an even 200mm for easy math in the future. If I set the Z value in the work coordinates to Z = -198.5 and presuming the clearance plane is also Z=0, the head shifts up by the length of the tool and the length of the tool holder combined, which places the tool tip exactly on the surface of the table. If I then subtract the thickness of the material from the 198.5 value, the tool tip then is set at the top of the stock/part. If the part/stock is 50mm thick, the Z value of the work offset should be Z=-148.5. If there is a clearance plane above the part, the tool will adjust up to that point in relation to the stock, which is Z=0 at the top.

Essentially, if I'm doing a job and want to simulate it, I just take my table to spindle nut difference and subtract the stock thickness so long as the stock is the thickness of the part, then make that value a negative Z value and the stock is properly placed exactly on top of the table. Here's a short video showing the simulation of a quick test shape, with showing the tool holder relationship between the Bobcad generated holder and the STL holder at the end. I think this gets me pretty close to where I need to be, though adjusting the Z value to 200mm would just make life a lot easier than 198.5mm, and really is no more or less valid that way. In fact, for CNC routing you are constantly changing that Z value by planing the wasteboard down, so there is never going to be a time where the machine and the simulation are identical. The relationships between the components will be just as accurate though, so the collision checking and visual simulated parts will be just as valid.

Bobcad V26 Simulation Test - YouTube