Scanning sharp edges with a Touch Trigger probe

[Chally72]

For those of you who have more in-depth scanning experience than I:

Any tips or tricks to improve scanning of sharp edges with a touch trigger probe? I am stuck doing something with a Touch Trigger, and I'm trying to bump the point density way up to try and get it to follow around the corner, but it's tough. Other than creating two scans and scanning towards the corner from each side, is there anything else I can do?

[Matthew D. Hoedeman]

For those of you who have more in-depth scanning experience than I:

Any tips or tricks to improve scanning of sharp edges with a touch trigger probe? I am stuck doing something with a Touch Trigger, and I'm trying to bump the point density way up to try and get it to follow around the corner, but it's tough. Other than creating two scans and scanning towards the corner from each side, is there anything else I can do?

IF you are doing this with cad data.....

Have you tried the EDGE POINT for the scan feature type? Generate the scan on the cad data, using edge points (w/ a surface sample hit to stay on the edge) and use a variable density scan.

[Bill **********]

Cut your PREHIT and RETRACT WAY down in that area. It should help you get in the sharp corners, but you will NEVER be able to really get into them. PCDmis calculates the approach vector of the next hit based on the last hits, so a square or less than 90 degree corner really isn't scannable. As long as there is some radius in the corner you can get a pretty good scan, but cutting your prehit and retract way down is about the only help that there is. This also speeds up your scans as you aren't spending as much hang time in the air.

Bill

[Chally72]

Thanks for the suggestions. I will try edge point in combination with scans up to the edge. Very little experience using that autofeature.

On a similar topic, how about a knife edge? Edge point might not do as well because it is chamfered irregularly from either side of a sheet metal 'blade.' I have parts that locate on this knife edge and need to find where they contact. Any suggestions?

[Matthew D. Hoedeman]

Thanks for the suggestions. I will try edge point in combination with scans up to the edge. Very little experience using that autofeature.

On a similar topic, how about a knife edge? Edge point might not do as well because it is chamfered irregularly from either side of a sheet metal 'blade.' I have parts that locate on this knife edge and need to find where they contact. Any suggestions?

2 vector points on each side, then 3-D lines, then intersect the 3-D lines maybe?

[Chally72]

How would this work, Matt?

Just to be sure we are on the same page, picture a knife. I have a part that rests on the cutting edge of the blade, perpendicular to the blade. It is a round part, so I need to find this cutting edge as a 'line' from handle to tip, (Yes, a weird knife with a totally straight blade,) and then make an intersection on this line later on. The blade edge is chamfered irregularly.

My thoughts were to take points on either side, then align to them and start a scan that took hits towards the cutting edge, crossing over the narrow part of the blade. The scan would start and end at slightly less than a ball radius from the edges of the blade. Do this two or three places on the blade, find high points from scans, and construct the 3D line from these high points. Only thing I could come up with :/

[Matthew D. Hoedeman]

How would this work, Matt?

Just to be sure we are on the same page, picture a knife. I have a part that rests on the cutting edge of the blade, perpendicular to the blade. It is a round part, so I need to find this cutting edge as a 'line' from handle to tip, (Yes, a weird knife with a totally straight blade,) and then make an intersection on this line later on. The blade edge is chamfered irregularly.

My thoughts were to take points on either side, then align to them and start a scan that took hits towards the cutting edge, crossing over the narrow part of the blade. The scan would start and end at slightly less than a ball radius from the edges of the blade. Do this two or three places on the blade, find high points from scans, and construct the 3D line from these high points. Only thing I could come up with :/

Hmmmm...... maybe this will help.

picture may be a little hard to see.....

BLACK = part
BLUE = measured points & constructed lines
RED = intersections of constructed line

[inspector212]

Try using a shank probe.

[Wingman]

Hmmmm...... maybe this will help.

picture may be a little hard to see.....

BLACK = part
BLUE = measured points & constructed lines
RED = intersections of constructed line


[ATTACH]5359[/ATTACH]

Doesn't ANGLE POINT in PCDMIS do this?

[Lee Johnson]

Be careful checking Knife edge parts with a TTP. The force may actually put a small divot on the edge. I had to go to using a tolmakers Microscope to check some Stainless Steel parts because the probe damaged them. I was measuring to the side of the ball so the force was less than 6 grams. The angle was very shallow and the parts were to a near perfect sharp edge.

[Matthew D. Hoedeman]

Doesn't ANGLE POINT in PCDMIS do this?

I don't know, I've never used it.

[Chally72]

Thank you, Lee, for the tip. The blade is stout enough that I don't think it will hurt it, but I will definitely examine it afterwards.

Wingman, I have never used this autofeature either. I will try today.

Don't want to use a shank probe, if I can avoid it. Causes more problems than it solves.

Matt, I see what you meant. That would work fine for that style blade, with a consistent, large angle leading towards the knife edge. I have attached a picture to show what the shape is of the blade I have to measure on this fixture. The knife edge was made on a belt sander, making the chamfer/angle small and inconsistent. Because of this, I'm sure angle point probably won't help either.

Edit: My drawing skills in MS paint are not formidable at 7 in the morning. Be warned that this picture is VERY bad!

[Matthew D. Hoedeman]

Thank you, Lee, for the tip. The blade is stout enough that I don't think it will hurt it, but I will definitely examine it afterwards.

Wingman, I have never used this autofeature either. I will try today.

Don't want to use a shank probe, if I can avoid it. Causes more problems than it solves.

Matt, I see what you meant. That would work fine for that style blade, with a consistent, large angle leading towards the knife edge. I have attached a picture to show what the shape is of the blade I have to measure on this fixture. The knife edge was made on a belt sander, making the chamfer/angle small and inconsistent. Because of this, I'm sure angle point probably won't help either.

Edit: My drawing skills in MS paint are not formidable at 7 in the morning. Be warned that this picture is VERY bad!

Is it possible for you to set it up in such a way that you could use "datum simulators"? Meaning, can you "fab up" something for you to push the taper point up against (like 2 pins) thus giving you something "good" to measure that you can push the "yucky" thing up against?

[Chally72]

I can't think of anything I could do. There's actually 4 of these things sticking out at compound angles in space to locate this part. I have no idea how the toolmaker even set the height of these blades in the holes. Maybe....These blades are sanded out at the top of dowel pins. A sleeve to slip over these blades, with a wider contact surface a known distance up from a stop. The sleeve would locate square off of the pin, farther down the shank. I'll check the squareness of the blade to the pin first, then use the sleeve to locate the blade surface. Boy, talk about getting ya dizzy just thinking about it.