4x cone diameters as datum

question for the pros out there
i have this casted part that has a datum target of 4x diameters.
these diameters are in a conical hole, and they re boxed 8.2mm down in the cone hole.

to get this datum
do i measure 4x circles 8.2mm down in the hole, and bestfit to them
or do i measure 4x cones, construct 4x circles using cone feature and 8.2mm down, and bestfit to them

ty all
love u
boy hol

Attached Files

mrollings53 over 2 years ago

I'm assuming Datum A is the surface these holes are normal to? Level to the surface, measure 4 circles at 8.2mm deep, rotate and origin to group of circles. Measure the rest of the hole features as cones then construct circles to measure the diameters, angles, and intersections of the cones.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
BKulpa over 2 years ago

I'd measure the cone and construct the circles at the 8.2 depth. Otherwise you'll have some cosine error from the draft.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Mike Ruff over 2 years ago

As BKulpa stated, you need to measure these as cones. If you probe a circle at a depth of 8.2, your diameter will measure .052mm smaller than it actually is at that depth (assuming perfect form) when using a Ø1.5mm probe. The error will increase with a larger tip and decrease with a smaller tip, but the error will always be there. See picture below for clarification. This happens because a circle is probed at 90°, while your surface is at 75°, so you are not probing normal to the surface. You are actually probing below the intended depth of the circle (.194mm below the specified 8.2mm when using a Ø1.5mm probe).

HTH
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
mrollings53 over 2 years ago

Glad I was corrected. Never gave a thought to the error in the size you would get.

As far as use as a Datum. The cosine error only matters if MMB is used in a FCF, correct?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
pcdmisstudent over 2 years ago

could i do this?
because i dont really understand how bestfit alignment works, so i rather use something my noob student brain can understand...

level to plane (A)
measure 4 cones and construct 4 circles
rotate between two circles
construct hole circle from 4 circles,
orgin to that circle.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Tater over 2 years ago

The 4 cones establish the -B- Datum. What sets -A- & -C- ? using the diameters at the required depth, measure all the cones then construct lines and mid-points between them. Then construct lines from the mid-points. Use these lines for alignment and set origin at intersection. this will split any error evenly.

HTH
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
pcdmisstudent over 2 years ago

Tater
datum A is plane parallell to cones, datum B is cones, datum C is plane pararllell to A
i didnt understand ur post
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Tater over 2 years ago

OK, I'll try again.

Looking at part from -Z- plus plane.
Measure the -A- Datum and level/origin it.
Measure the 4 cones at 8.2 depth.
Use the cones measurement to construct 4 lines along the perpendicular axis ( should make a square).
Use the lines to construct 4 Mid-points .
Use the 4 Mid-points to construct 2 more lines ( 1 X , 1 Y).
These 2 lines should intersect at the center of the 4 cones.
Align using 1 of the 2 Mid-point lines.
Origin X&Y at intersect point of 2 Mid-point lines.

HTH
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
pcdmisstudent over 2 years ago in reply to Tater

Ty now i understand =)
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
pcdmisstudent over 2 years ago

Tater
advanced noob question from me
is this method kinda like just doing the bestfit least squared method?
or whats the different
NinjaBadger
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel