Parametric Programming?

[benj118]

Anyone have experience with parametric programming and the expected efficiency if any?

[slesholdofthreep]

This a term usually associated w/ cnc g-code. Are you talking about PCDMIS variables? Depending on what you need sometimes they are your only option.

[benj118]

This a term usually associated w/ cnc g-code. Are you talking about PCDMIS variables? Depending on what you need sometimes they are your only option.

Parametric as in one program to run a similar family of parts picking variables from a spread sheet.

[craiger_ny]

Yes it can be done but your question can't really be answered as it depends heavily on part configuration. J has had some issues with movepoints, you'll have to do some extra coding there to work around that. But is is completely doable and it is likely you will bennefit from it but like I say it is part configuration dependant.

Do you plan on storing your variables in a sreadsheet, database, or text file? If you want PCDMIS to pull direct text is probably the way to go but if PCDMIS can read from a database I'd like to see it, that would be cool.

[cmmguy]

Anyone have experience with parametric programming and the expected efficiency if any?

It works ok, PCDMIS is inefficient at the task. As Craig pointed out, you will have to program your move points with variables because the move relative is weak in pcdmis.

You cant embed a front end, you will have to employ an external selection tool unless you want a simple "Enter 1 for partA, 2 for partB, etc..."

Because of the changing nominals in pcdmis, I use alot of variables in my programs for nominals.

[jon_dav_uk]

We have a huge parametric program in MM4 which at some stage I need to convert to the demon.

I have not had a chance to really look into it but initial thoughts were to have a custom application which does all the "clever" stuff and just having the demon to read in the required data via text file, assigning to variables and running. This will work but more than likely require a fair bit of effort, will be worth it in the end tho.

I would be interested in seeing what you come up with............

[craiger_ny]

Because of the changing nominals in pcdmis, I use alot of variables in my programs for nominals.

I hate that. I have ran into many cases of variables turning into literals not only in features and dimensions but in comments as well. It drives me nuts. Alas though we don't really need code, let alone code that works. Just buttons and cad models should do.

From what I have heard from others MM4 has the leg up on parametric programming. But MM4 is code intense and code is so yesterday.

[jon_dav_uk]

Agree 100% with you there craiger, MM4 is as robust as it gets, no bugs, never falls over, great for parametrics, super fast but who the **** understands Unix....certainly not I.

Cant wait to start converting it to the demon, crash, bang, wallop springs to mind

[JamesMannes]

We have a huge parametric program in MM4 which at some stage I need to convert to the demon.

I have not had a chance to really look into it but initial thoughts were to have a custom application which does all the "clever" stuff and just having the demon to read in the required data via text file, assigning to variables and running. This will work but more than likely require a fair bit of effort, will be worth it in the end tho.

I would be interested in seeing what you come up with............

FYI there is an MM4 module that you can buy for DMIS and it will run your MM$ programs. Apparently when you open this module it opens a new window that looks just like MM4 so you in essence run MM4 through PC-DMIS.

[cmmguy]

FYI there is an MM4 module that you can buy for DMIS and it will run your MM$ programs. Apparently when you open this module it opens a new window that looks just like MM4 so you in essence run MM4 through PC-DMIS.

Doesnt it just shell mm4??

[jon_dav_uk]

Yeah we have the MM4 emulator which runs ok which is why I have not had an urgent request to convert to PC Dmis.

At some stage though we need to convert across as the additional functionality / flexibility and better reports in PC Dmis etc will be a big help. I also have a horrible feeling that at some stage, the whole MM4 thing will simply stop working (maybe I am just a pessimist) and with no support anymore, this would leave us deep in the ****.

EDIT: - Yep it does just shell MM4 cmmguy

[dwade]

Actually there is still some un-official MM4 support from B&S. Some of the older AE still help with the issues but not on a regular basis. We still use MM4 supplied programs and I had a problemarise that I could not fix and the customer's programmer had no idea eother so I sent the program to B&S and they found the issue within a day.

[jon_dav_uk]

Thats good to know, I have a few contacts at Hexagon who are ex-MM4 guru's who I have always suspected might help out but nothing official

[benj118]

Thanks for all of the replies, I have added some of the code for the start of a basic parametric program below as an example. I guess what I am really after if anyone has used and programmed with parametric programming and any advantage you found? (Some of the code below was written with scanning with an SP600 head)




RECALL/ALIGNMENT,EXTERNAL,a6_sm,A6LINK FIXT:a6_sm
LOADPROBE/PROBE6
TIP/T1A0B0, SHANKIJK=0.005, -1, 0.021, ANGLE=-0.274
MOVE/POINT,NORMAL,-33,-6,178
ASSIGN/PART_NUMBER = 6I9239
CS4 =SCRIPT/FILENAME= C:\CMM\PROGRAMS_ PROVEN\LINK_SCANNED\MACHINED_SCAN.BAS
FUNCTION/CallExcel,SHOW=YES,,
STARTSCRIPT/
ENDSCRIPT/
ASSIGN/H1X = -30.5
ASSIGN/H1Y = -1
ASSIGN/H1Z = 139.9
ASSIGN/H1I = 0
ASSIGN/H1J = 0
ASSIGN/H1K = 1
ASSIGN/H1DIA = 23.22
H1 =AUTO/CIRCLE,SHOWALLPARAMS = YES,SHOWHITS = NO
THEO/H1X,H1Y,H1Z,H1I,H1J,H1K,H1DIA
ACTL/-32.423,-2.254,139.655,0,0,1,23.201
TARG/H1X,H1Y,H1Z,H1I,H1J,H1K
THEO_THICKNESS = 0,RECT,IN,STRAIGHT,LEAST_SQR,ONERROR = NO,$
AUTO MOVE = BOTH,DISTANCE = 10,RMEAS = None,None,None,$
READ POS = NO,FIND HOLE = CENTER,REMEASURE = NO,$
NUMHITS = 3,INIT = 1,PERM = 1,SPACER = 5,PITCH = 0,$
START ANG = 0,END ANG = 0,DEPTH = 5,$
VOID DETECT = NO,$
ANGLE VEC = 0.9999858,0,-0.0053253
XYZ_ORIGIN =ALIGNMENT/START,RECALL:A6LINK FIXT:a6_sm,LIST=YES
ALIGNMENT/TRANS,XAXIS,H1
ALIGNMENT/TRANS,YAXIS,H1
ALIGNMENT/TRANS,ZAXIS,H1
ALIGNMENT/END
ASSIGN/H2X = 63
ASSIGN/H2Y = -29
ASSIGN/H2Z = 0
ASSIGN/H2I = 0
ASSIGN/H2J = 0
ASSIGN/H2K = 1
ASSIGN/H2DIA = 23.22
H2 =AUTO/CIRCLE,SHOWALLPARAMS = YES,SHOWHITS = NO
THEO/H2X,H2Y,H2Z,H2I,H2J,H2K,H2DIA
ACTL/63.153,-28.849,0,0,0,1,23.187
TARG/H2X,H2Y,H2Z,H2I,H2J,H2K
THEO_THICKNESS = 0,RECT,IN,STRAIGHT,LEAST_SQR,ONERROR = NO,$
AUTO MOVE = BOTH,DISTANCE = 10,RMEAS = None,None,None,$
READ POS = NO,FIND HOLE = DISABLED,REMEASURE = NO,$
NUMHITS = 5,INIT = 3,PERM = 0,SPACER = 5,PITCH = 0,$
START ANG = 0,END ANG = 0,DEPTH = 5,$
VOID DETECT = NO,$
ANGLE VEC = 0.9999858,0,-0.0053253
SHFACEPNT1 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H2X+(H2DIA*.7),H2Y,H2Z,H2I,H2J,H2K
ACTL/79.253,-29.001,-0.059,0,0,1
TARG/H2X+(H2DIA*.7),H2Y,H2Z,H2I,H2J,H2K
SHFACEPNT2 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H2X,H2Y+(H2DIA*.7),H2Z,H2I,H2J,H2K
ACTL/62.999,-12.745,-0.124,0,0,1
TARG/H2X,H2Y+(H2DIA*.7),H2Z,H2I,H2J,H2K
SHFACEPNT3 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H2X-(H2DIA*.7),H2Y,H2Z,H2I,H2J,H2K
ACTL/46.746,-29.002,0.09,0,0,1
TARG/H2X-(H2DIA*.7),H2Y,H2Z,H2I,H2J,H2K
SHFACEPNT4 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H2X,H2Y-(H2DIA*.7),H2Z,H2I,H2J,H2K
ACTL/62.998,-45.253,0.151,0,0,1
TARG/H2X,H2Y-(H2DIA*.7),H2Z,H2I,H2J,H2K
SHFACEPNT5 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H1.X+(H1DIA*.7),H1.Y,H1.Z,H1I,H1J,H1K
ACTL/16.252,0.004,0,0,0,1
TARG/H1.X+(H1DIA*.7),H1.Y,H1.Z,H1I,H1J,H1K
SHFACEPNT6 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H1.X,(H1.Y+H1DIA*.7),H2.Z,H1I,H1J,H1K
ACTL/-0.001,16.255,-0.059,0,0,1
TARG/H1.X,(H1.Y+H1DIA*.7),H2.Z,H1I,H1J,H1K
SHFACEPNT7 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H1.X-(H1DIA*.7),H1.Y,H1.Z,H1I,H1J,H1K
ACTL/-16.258,-0.002,0.155,0,0,1
TARG/H1.X-(H1DIA*.7),H1.Y,H1.Z,H1I,H1J,H1K
SHFACEPNT8 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/H1.X,H1.Y-(H1DIA*.7),H1.Z,H1I,H1J,H1K
ACTL/-0.002,-16.256,0.208,0,0,1
TARG/H1.X,H1.Y-(H1DIA*.7),H1.Z,H1I,H1J,H1K
SHOE_FACE_PLN=FEAT/PLANE,RECT,TRIANGLE
THEO/31.5,-14.5,0,0,0,1
ACTL/31.498,-14.5,0.045,0.004779,0.0082854,0.9999543
CONSTR/PLANE,BF,SHFACEPNT1,SHFACEPNT2,SHFACEPNT3,SHFACEPN T4,SHFACEPNT5,SHFACEPNT6,SHFACEPNT7,SHFACEPNT8,,
OUTLIER_REMOVAL/OFF,3
FILTER/OFF,WAVELENGTH=0
ASSIGN/LIN_OS1 = 0
ASSIGN/LIN_OS2 = 29
LIN1 =FEAT/LINE,RECT,UNBND
THEO/0,0,0,1,0,0
ACTL/0,0,0,0.9999972,0.0023851,0
CONSTR/LINE,OFFSET
SURFACE NORMAL = 0,0,1,MULTI POINT
ID = H1,H2,,
OFFSET = LIN_OS1,LIN_OS2
TEMP_LOCATION=ALIGNMENT/START,RECALL:XYZ_ORIGIN,LIST=YES
ALIGNMENT/LEVEL,ZPLUS,SHOE_FACE_PLN
ALIGNMENT/TRANS,ZAXIS,SHOE_FACE_PLN
ALIGNMENT/TRANS,XAXIS,H1
ALIGNMENT/TRANS,YAXIS,H1
ALIGNMENT/ROTATE,XPLUS,TO,LIN1,ABOUT,ZPLUS
ALIGNMENT/END
MOVE/POINT,NORMAL,H1.X,H1.Y,H1.Z+100
TIP/T1A90B180, SHANKIJK=-0.002, -1, -0.022, ANGLE=-179.776
MOVE/POINT,NORMAL,H1.X,H1.Y-100,H1.Z
PREHIT/ 6
RETRACT/ 6
ASSIGN/X1_X = -58
ASSIGN/X1_Y = -15.96
ASSIGN/X1_Z = -11
ASSIGN/X1_I = 0
ASSIGN/X1_J = -1
ASSIGN/X1_K = 0
X1 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/X1_X,X1_Y,X1_Z,X1_I,X1_J,X1_K
ACTL/-57.993,-16.125,-11.005,0,-1,0
TARG/X1_X,X1_Y,X1_Z,X1_I,X1_J,X1_K
ASSIGN/X2_X = -52.3
ASSIGN/X2_Y = -23.46
ASSIGN/X2_Z = -126.8
ASSIGN/X2_I = 0
ASSIGN/X2_J = -1
ASSIGN/X2_K = 0
X2 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/X2_X,X2_Y,X2_Z,X2_I,X2_J,X2_K
ACTL/-52.293,-23.387,-126.801,0,-1,0
TARG/X2_X,X2_Y,X2_Z,X2_I,X2_J,X2_K
ASSIGN/X3_X = 188
ASSIGN/X3_Y = -53.46
ASSIGN/X3_Z = -58.5
ASSIGN/X3_I = 0
ASSIGN/X3_J = -1
ASSIGN/X3_K = 0
MOVE/POINT,NORMAL,X2_X,(X2_Y-25),X2_Z
MOVE/POINT,NORMAL,-77,-169,-54
MOVE/POINT,NORMAL,183,-170,-55
ASSIGN/X3_X = 188
ASSIGN/X3_Y = -53.46
ASSIGN/X3_Z = -58.5
ASSIGN/X3_I = 0
ASSIGN/X3_J = -1
ASSIGN/X3_K = 0
X3 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/X3_X,X3_Y,X3_Z,X3_I,X3_J,X3_K
ACTL/188.019,-53.328,-58.497,0,-1,0
TARG/X3_X,X3_Y,X3_Z,X3_I,X3_J,X3_K
MOVE/POINT,NORMAL,190,-115,169
TIP/T1A90B0, SHANKIJK=0.002, 1, 0.022, ANGLE=179.776
MOVE/POINT,NORMAL,189.749,45.653,57.145
MOVE/POINT,NORMAL,190.285,46.69,-62.432
ASSIGN/A3_X = 188
ASSIGN/A3_Y = -17.46
ASSIGN/A3_Z = -58.5
ASSIGN/A3_I = 0
ASSIGN/A3_J = 1
ASSIGN/A3_K = 0
ASSIGN/A3_OFF =
A3 =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
THEO/A3_X,A3_Y,A3_Z,A3_I,A3_J,A3_K
ACTL/187.998,-17.128,-58.504,0,1,0
TARG/A3_X,A3_Y,A3_Z,A3_I,A3_J,A3_K
ASSIGN/X1OFF = 0
ASSIGN/X2OFF = -7.5
ASSIGN/X3OFF = -37.5
DAT_X_PLN =FEAT/PLANE,RECT,TRIANGLE
THEO/25.9,-15.96,-65.433,0,-1,0
ACTL/25.898,-15.946,-65.404,0.0008184,-0.9999976,-0.0020152
CONSTR/PLANE,OFFSET
ID = X1,X2,X3,,
OFFSET = X1OFF,X2OFF,X3OFF
DATDEF/FEATURE=DAT_X_PLN,DATUM=X

[cmmguy]

Sometimes PCDMIS doesnt recognise a variable if you just cut and paste. if you have a sitch that doesnt seem right, check the values of the variables.

You are on the right track. I usually assign all the variables at the beginning of the program. It seems easier to manage feature rearrangement. less errors.