Find Hole Woes

[#2#]

Hello Everybody

I am having some trouble with findhole in the autocircle window. I brought this up before and never got a satisfactory answer.

Lets say I have a simple part my part is actually much more complicated. A plate with two holes in it. I take a manual plane, a manual line on one edge and a manual circle on one hole. I level the plane, rotate to the line, set the plane to z orgin and set the hole to the xy orgin. Now, I want to measure the other hole which is supposed to be at 1,0,0 and I want to measure it .05 below the surface. I plug all this into the autocircle window and select findhole b/c my hole will float around a bit and I don't want to have to do anything manual at this point.

Now, to the problem. My hole is not perp. to the plane. As the probe enters the hole it "searches" for the hole (moving in Z) until it is down the check distance past nominal. Because the hole is not perp. the probe hits before it reaches the max. check distance. It then backs up and searches as per normal. The next "search" it does not hit anything but for some reason it searches the check distance past where it hit before. Then, when it measures the hole it measures .05 below where it hit before.

I know I could change my check distance but that is only a workaround.

3.7mr3 - didn't have this problem before installing mr3.

[justncredible]

Umm, I think if I remeber class right is the way find hole works, it repetes the pattern to make sure then goes to measure it. I never used it, sorry.......

[#2#]

Umm, I think if I remeber class right is the way find hole works, it repetes the pattern to make sure then goes to measure it. I never used it, sorry.......

Yep that is the way it works but it changes the z height where it measures the hole. It shouldn't do that. It should only be changing x and y.

[John Riggins]

Have you set the vectors for the non-perp hole?

[#2#]

No, the hole is supposed to be perp. to the plane but due to machining error it is not.

[justncredible]

Ah are you setting the zheight just for the find hole?

After it finds the hole then it may switch to the depth in autofeature.

[Dannyrl]

Do you have read position selected? If you do, uncheck it. Not sure but I think it may be messing you up.

[Dannyrl]

sorry. I was wrong read position doesn't seem to affect the check

[#2#]

Ah are you setting the zheight just for the find hole?

After it finds the hole then it may switch to the depth in autofeature.

I'm setting the Z height to the point where the hole intersects the plane (0). Then, I am setting the depth to the depth where I want the hits taken (.05). The problem seems to be that as it goes it changes the z nominal to wherever it last encountered a hit. So, if I have it originally set at Z=0 and it searches and encounters the wall at say Z=-.03 the next search it uses Z=-.03 as the nominal. Then, the next search it starts at Z=-.03 and searches the set depth below that. If it hits again at Z=-.05 it will use that as the nominal for the next search. Then, if it "finds" the hole it will take the hits for the circle at Z=-.1 (-.05 nominal -.05 depth).

[bw_bob]

Have you tried using initial,permanent hits & setting you depth in the hits drop down column.

[justncredible]

I think I understand what it is doing then, once it finds the hole depth it is thinking that depth is the top of the hole and then from that point goes to the .05 you have set. You are searching in the Z, try zero Z, so when it finds the hole it does not reset the top of the part. Keep the feature depth at -.05 and see if that works????????

[kingsld1]

I have a routine that I've been using for when I have problems with FIND HOLE. I get large sheet metal parts with small diameters and large tolerances that wander all over the place and this routine always works for me.

The XYZIJK dia, pitch, # of hits etc. is stored as variables. The parameters in the autocircle use the variables to define the hole.
If FIND HOLE fails then the program asks you if you want to measure the hole and surface manually. After you measure the hole and surface, the XYZ data is assigned to variables and sent back to the autocircle routine.

Things I like about the routine is that the stats package is unaffected. Data is always sent to the correct feature name. The program doesn't have to be re-started in mid stream. You don't have to worry about editing the program for one oddball part. It keeps other people out of the program! Some manual hits need to be taken but you're right back in the program taking the same # of hits at the desired depth.

BTW this was written in ver. 3.25. It is currently setup for a hole in -X. The SKIP when you click NO on the dialog box has not been programmed in yet. (I've never needed it so I've never gotten around to fixing it.)


Hope this helps!

DRK



COMMENT/DOC,Use when the built-in "FIND HOLE" command won't work consistently.
,
,Added I,J,K,DIA,HITS and PTCH for this version.
,This is setup for a -X hole.
,
,
,
ONERROR/UNEXPECTED_HIT,OFF
ONERROR/UNEXPECTED_HIT,GOTO CIR33_ERR_HANDLER
CHECK/ 1,0.75
ASSIGN/X = 6.4
ASSIGN/Y = 58.49
ASSIGN/Z = -3.2
ASSIGN/I = -1
ASSIGN/J = 0
ASSIGN/K = 0
ASSIGN/DIA = 1.3
ASSIGN/HITS = 8
ASSIGN/PTCH = 0.4
CIR33_AGAIN=LABEL/
MOVE/CLEARPLANE
CIR33 =AUTO/CIRCLE, SHOWALLPARAMS=YES, SHOWHITS=NO
THEO/X,Y,Z,I,J,K,DIA
ACTL/6.4,58.49,-3.2,-1,0,0,1.597
TARG/X,Y,Z,I,J,K
THEO_THICKNESS = 0, RECT, IN, STRAIGHT, LEAST_SQR, ONERROR = NO ,$
AUTO MOVE = NO, DISTANCE = 0, RMEAS = None, READ POS = NO, FIND HOLE = CENTER, REMEASURE = NO ,$
NUMHITS = HITS, INIT = 3, PERM = 0, SPACER = 1, PITCH = PTCH ,$
START ANG = 0, END ANG = 360, DEPTH = 0.75 ,$
ANGLE VEC = 0.0005507,0.0006377,0.9999996
GOTO/CIR33 _CONTINUE

CIR33_ERR_HANDLER=LABEL/
MODE/MANUAL
C3 =COMMENT/YESNO,Measure CIR33 manually?
,
,Click YES to MEASURE
,Click NO to SKIP FEATURE
,Click CANCEL to END PROGRAM.
CIR33_MAN =FEAT/CIRCLE,RECT,IN,LEAST_SQR
THEO/X,Y,Z,I,J,K,DIA
ACTL/6.4,58.49,-3.2,-1,0,0,1.3
MEAS/CIRCLE,4,WORKPLANE
MOVE/CLEARPLANE
HIT/BASIC,6.398,58.861,-2.672,0,-0.5747966,-0.8182963,125.891,-26.481,-663.71
HIT/BASIC,6.4,58.859,-3.728,0,-0.573418,0.819263,125.897,-26.483,-665.25
HIT/BASIC,6.401,57.845,-3.296,0,0.9891617,0.1468303,125.896,-25.977,-664.504
HIT/BASIC,6.401,59.141,-3.296,0,-0.9892494,0.1462379,125.898,-27.717,-664.504
ENDMEAS/
COMMENT/OPER,Measure point on surface
COMMENT/DOC,Surface Measurement is in case the surface isn't where it should be.
CIR33_SURF =FEAT/POINT,RECT
THEO/X,Y,Z,I,J,K
ACTL/6.4,58.49,-3.2,-1,0,0
MEAS/POINT,1
MOVE/CLEARPLANE
HIT/BASIC,6.4,58.49,-3.2,-1,0,0,124.689,-23.586,-664.486
ENDMEAS/
ASSIGN/X = CIR33_SURF.X
ASSIGN/Y = CIR33_MAN.Y
ASSIGN/Z = CIR33_MAN.Z
ONERROR/UNEXPECTED_HIT,OFF
COMMENT/OPER,Hit OK to go back to DCC movement.
MODE/DCC
GOTO/CIR33_AGAIN
CIR33 _CONTINUE=LABEL/

ONERROR/UNEXPECTED_HIT,OFF
ONERROR/UNEXPECTED_HIT,GOTO NEXTHOLE_ERR_HANDLER
NEXTHOLE_ERR_HANDLER=LABEL/

[Dannyrl]

Goodluck,
I understand your problem now. I measure a lot of extruded holes in sheet metal and the area around the hole gets sucked in kind of like a countersink. I have the same problem .When the last search hit is near the hole, which is really a couple millimeters past the surface, it thinks that last hit is the surface. I haven't found the fix yet. When I have that problem the only thing I could come up with is increasing the prehit and retract. After it finds the hole and finds the surface it will measure at the depth you want. I usually have the initial a nd permenant hits on 3.

[#2#]

Well, the only thing I have come up with is to double measure the hole. Use the find hole to measure the hole (it will measure at the wrong depth), then align to the hole and measure another hole with the nominals set to 0,0,-.05.

[Dannyrl]

If I find anything that works better I'll let you know