Profile

[[Ni...]]

Hi,
Need help getting the profile to a dimension. I am very new to the curve feature and was wondering if anyone could help.
I scan the bottom side of that feature as a 3d curve.
Then i recall that feature using free form surface.
Then i create the profile characteristic and choose the free form surface which recalls the 3d curve. I choose Datum A as the datum.

I get around .020.
I am not sure which option to choose in the bilateral section. Or would it even mater? Hopefully someone can help with this. Zeiss created a program that checks this dimension and it comes out within spec on their program. Just wondering what i am doing wrong.

Thank You

[[Jo...]]

I have curve , but haven't had the course yet. What is the biggest aid for me is turn on CAD evaluation. You can then select the profile and see what is going on. If it is just one point or filtering etc.

[[Ch...]]

Hi.

Couple things:
The way you have done it sounds like 1 correct method.
This DRF (Datum A only) I don't think this fully constrains this profile, but anyhow...
"-A- only DRF" does allow rotational best fit (and perhaps translation - but DRF does not appear complete.)
Technically if the profile is A<---->B The (R.08) should probably be basic.
Feel free to share this info with Eng Dept and ask them for a second look.
Are you measuring this a 3D surface 360° around the part? I'm not sure if we're looking at a cutout view or the entire actual side view of the part.
If Zeiss programmed this, I assume it's mostly correct. 3D curve is considered the 'black arts' of CMM. Be sure to copy their scan settings and most importantly all evaluation settings. I would try to copy their feature from the other program directly if possible.
Also, if recalling into FFS, you can also just slice the CAD model and create a FFS and bypass the 3D curve. With some features, it makes it easier to define scan path by using 3D Curve first.
As a sanity check, you could measure the 2 planes (and the R.08 if feasible), and the max value of those 2 shouldn't be much more than what you get with FFS.

As far as the bilateral-one result. Yes that is correct evaluation method.


Good luck!

[[Ja...]]

A common mistake made when learning curve is the number of measured points to number of evaluation points. When you create your curve in CAD-Create Features, The Default number of points is 100. If you create a curve with 100 points, then you want to have at least 3 times that many points in the the strategy of the curve. That is the first thing I would check.

[[Ro...]]

Is there a reason you are using curve to create a Free Form?
Why not just create your scans using free form by itself?