Quindos to Dmis

[mfzaring]

Brand new to PC-DMIS, programmed in quindos for the last decade. Wanted to see if there was anyone else that has made the change and could give me some advise or tips. I've had the software on my computer for the last couple of days just messing around with a model. I guess one of the things I want to know is; is there a way to manipulate points like in quindos? I want to scan some features and then select points from those scans, find the extreme points in a specific axis then get a width. Any help would be appreciated.

[AndersI]

Which version of PC-DMIS? 2019 R1 has a new construction, "tertiary datum", which can get you the 'high point' in any direction.

There are also constructions for line and circle from "Scan segment" which may be useful for you, depending on a lot of things you didn't tell us. This exists in all(?) versions of PC-DMIS.

[DAN_M]

I have programmed PC DMIS for the last decade but will be taking my first Quindos class in Rhode Island very soon. Feel free to reach out any time, we could probably help each other out a lot.

Yes. You can scan the part then use some expressions/math to extract the MIN/MAX points from that scan. Then construct them and do stuff with them. I just have never done it before.

JEFMAN is literally a genius and will know.

[JEFMAN]

DAN_M : thanx a lot, even if it's not true !
mfzaring : the direction is defined by a vector, that you can create from a feature (LIN1.IJK) or from an assignment (ASSIGN/V1=UNIT(MPOINT(3,2,5))
Here, MPOINT helps creating a coordinate variable, UNIT makes its length = 1.
Then, if you look at a width along V1, you can project each point of the scan on it with a dot product :
ASSIGN/V2=DOT(SCN1.HIT[1..SCN1.NUMHITS].XYZ,V1) which gives an array of values. Note here that the projection seems to be done from coordinates directly, but it's in reality the projection of n vectors between scan hits and origin along V1.
Then, you can search the min value with ASSIGN/V3=MIN(V2) and the max with ASSIGN/V4=MAX(V2)
You can also search the min point ASSIGN/V5=MININDEX(V2), then create a generic point with SCN1.HIT[V5].XYZ.


You can go here : ftp://ftp.wilcoxassoc.com/docs/ then select your version, and dowload the chapter "Using expressions and variables".

[mfzaring]

Thanks guy, this is exactly what I was looking for. DAN_M feel free to reach out with any quindos questions, just a heads up tho...they are completely different lol

[DAN_M]

Thanks I appreciate it, will def keep you in mind.

[mfzaring]

JEFMAN below is posted the 2 scans I am trying to get the extreme points for, could you show me some code that would get the most -Y for the front scan and the most +Y for the back scan? I'm just not used to this format yet and having an example would help greatly. Thanks


FRONT_SCAN =FEAT/SCAN,LINEAROPEN,NUMBER OF HITS=13
,SHOW HITS=NO,SHOWALLPARAMS=NO
MEAS/SCAN
BASICSCAN/LINE,NUMBER OF HITS=13,SHOW HITS=NO,SHOWALLPARAMS=NO
ENDSCAN
ENDMEAS/
MOVE/INCREMENT,<0,0,5>
MOVE/INCREMENT,<0,6,0>
BACK_SCAN =FEAT/SCAN,LINEAROPEN,NUMBER OF HITS=15
,SHOW HITS=NO,SHOWALLPARAMS=NO
MEAS/SCAN
BASICSCAN/LINE,NUMBER OF HITS=15,SHOW HITS=NO,SHOWALLPARAMS=NO
ENDSCAN
ENDMEAS/

[mfzaring]

This is what I have so far, but not sure what to do with the points...(where do I use the variables to create something I can do a distance on) or how to make sure I am looking at the Y axis min and max. Pretty sure I need v2 and vv4

ASSIGN/V1=UNIT(MPOINT(3,2,5))
ASSIGN/V2=DOT(FRONT_SCAN.HIT[1..FRONT_SCAN.NUMHITS].XYZ,V1)
ASSIGN/V3=MIN(V2)
ASSIGN/V4=MAX(V2)
ASSIGN/VV1=UNIT(MPOINT(3,2,5))
ASSIGN/VV2=DOT(BACK_SCAN.HIT[1..BACK_SCAN.NUMHITS].XYZ,VV1)
ASSIGN/VV3=MIN(VV2)
ASSIGN/VV4=MAX(VV2)

[JEFMAN]

You're looking for distance along Y so you could :
ASSIGN/V1=MPOINT(0,1,0).................................. .................................vector Y+
ASSIGN/P1=(FRONT_SCAN.XYZ+BACK_SCAN.XYZ)/2..................approx middle point between scans
ASSIGN/V2=DOT(FRONT_SCAN.HIT[1..FRONT_SCAN.NUMHITS].XYZ-P1,V1)
ASSIGN/V3=MAXINDEX(V2)
ASSIGN/V4=FRONT_SCAN.HIT[V3].XYZ
ASSIGN/V5=FRONT_SCAN.HIT[V3].IJK
F1 =GENERIC/POINT,DEPENDENT,CARTESIAN,$
NOM/XYZ,<V4.X,V4.Y,V4.Z>,$
MEAS/XYZ,<V4.X,V4.Y,V4.Z>,$
NOM/IJK,<V5.I,V5.J,V5.K>,$
MEAS/IJK,<V5.I,V5.J,V5.K>

[mfzaring]

Thanks JEFMAN you have pointed me in the right direction, things are starting to make sense!

[JEFMAN]

You're welcome !