Surface profile from CAD model

[Underspec]

Using a CAD model to do a perimeter scan on a surface for profile. It's coming out bad

I have the settings set to:

Exec Controls - Defined
Nominals Method - FINDNOMS
Hit Controls - Vectors

Am I missing something? Shouldn't the CAD model give me the nominals? My part is aligned to the CAD model. I did not use CAD=Part and "find nominals" is not turned on in setup.

[Matthew D. Hoedeman]

NOMINAL METHOD = MASTER
EXECUTE = NORMAL (you do not want re-learn)

Once you select the boundaries, click on GENERATE (finds all the noms and vectors from the cad model) before you click on create (with measure un-marked). These are the commands/options from V3.7 MR3, you will have to figure out what they match in V4.1

[JamesMannes]

Using a CAD model to do a perimeter scan on a surface for profile. It's coming out bad

I have the settings set to:

Exec Controls - Defined
Nominals Method - FINDNOMS
Hit Controls - Vectors

Am I missing something? Shouldn't the CAD model give me the nominals? My part is aligned to the CAD model. I did not use CAD=Part and "find nominals" is not turned on in setup.

How bad is bad? Maybe it is bad? You did choose the CAD surfaces to include in your scan correct? I don't remember if multiple surfaces can be chosen in 3.7.

[JamesMannes]

NOMINAL METHOD = MASTER
EXECUTE = NORMAL (you do not want re-learn)

Once you select the boundaries, click on GENERATE (finds all the noms and vectors from the cad model) before you click on create (with measure un-marked). These are the commands/options from V3.7 MR3, you will have to figure out what they match in V4.1

Do not set to "master" unless you want that first piece you measure to be the nominal values from that scan on. When using CAd you should always "Nominals" or "Find Nominals". "Nominals" and "Find Nominals" do almost the same thing. Use "Find Nominals" when your piece may not be real close to nominal and use "Nominals" when you expect the piece to be relatively close to nominals. I know that's a screwy way to look at it but that is how Jonathan Dove (BNS scanning guru Elgin.Ill) explained it to me when he paid us a visit for some scanning issues with earlier versions.

As far as "Relearn" and "Normal" goes, if using an SP600 choose "Relearn" so as not to "stitch scan" your part. Use "Normal" if you want to "stitch scan" your part.

[Underspec]

When MASTER is selected from the Nominals list, PC-DMIS will consider the measured data obtained the first time the scan is being learned as nominal data. Subsequent executions of this scan will be compared with this measured data. When MASTER is used with a DCC scan and Normal Mode is selected from the Execute list, PC-DMIS will execute a Stitch-type scan using the measured data.

That's not what I want is it?

[Matthew D. Hoedeman]

Do not set to "master" unless you want that first piece you measure to be the nominal values from that scan on. When using CAd you should always "Nominlas" or "Find Nominals". "Nominals" and "Find Nominals" do almost the same thing. Use "Find Nominals" when your piece may not be real close to nominal and use "Nominals" when you expect the piece to be relatively close to nominals. I know that's a screwy way to look at it but that is how Jonathan Dove (BNS scanning guru Elgin.Ill) explained it to me when he paid us a visit for some scanning issues with ealier versions

That is taken care of when you follow my directions, click on GENERATE and CREATE without MEASURING. Then all your nominals are cad master data.

[JamesMannes]

That is taken care of when you follow my directions, click on GENERATE and CREATE without MEASURING. Then all your nominals are cad master data.

Ok, I'll accept that. Once again The Hoedeman is correct.

[Matthew D. Hoedeman]

Ok, I'll accept that. Once again The Hoedeman is correct.

No problem. Of course, those are directions for TTP, I have no idea what you would want for one of the continuous contact scanner heads.

[Underspec]

Alright tried what matt said. Now the surface profile is perfect (0.0000) Something isn't right.

[Matthew D. Hoedeman]

Alright tried what matt said. Now the surface profile is perfect (0.0000) Something isn't right.

Well, it WILL be perfect, until you execute it. AND, you may have to re-ID the dimension as well (re-type in the feature ID).

[Mike Johnston]

Correct workplane? Had to ask...... No insult intended

[Matthew D. Hoedeman]

Correct workplane? Had to ask...... No insult intended

Workplane has no effect on the "3-D" features, basically, anything in the CTRL-F menu (auto features) and anything in the DCC Scan menu.

[Underspec]

Yes I already executed it. And it's showing 0 deviation.

[Matthew D. Hoedeman]

Yes I already executed it. And it's showing 0 deviation.

Did you execute it (CTRL-Q) or did you execute it (CTRL-E)?

Did you try re-entering the feature ID in the dimension?

Did you try saving the program, close it, and re-open it?

Are you using the TTP or a continuous contact method?

(I already know you are a 4-head!)

[Underspec]

I used CTRL-E on the feature. And i set it to defined for continous contact. I have a SP600.