vectors, vectors, and more vectors

[Shelley Liebman]

I am working in the X-plus plane....and have these lovely holes in my part that are at goofy angles to the work plane.
I have one at a vector of .707, -.707, 0--so that it is pointing towards the right front corner of my machine.
I also have 3 others at goofy angles but lets start with this one.
I physically taught the machine where the hole was, and of course it asked me do I want to change the workplane, to which I answered no. I want to keep it in the Xplus.
Upon creating the hole, I then had to manually edit the vector of the hole itself, as the machine thinks the vector should be 1, 0, 0. I changed that to .707, -.707, 0. I also had to manually change the vector of each individual hit so that it takes the hits properly.
When I run the feature through, it takes clean hits, but the actuals are still giving me a vector of 1, 0, 0.

Any ideas of what I am doing wrong?????????

[tking]

Just a thought, can you probe a cylinder? Three levels, 3 or 4 points at each level. It'll nail the vector and not care about the working plane.

TK

[Matthew D. Hoedeman]

Well, try this:
1) Get the vector set to what you want it to be in the edit window
2) place the curson ON the feature and hit F9
3) A window will pop up with the 'info' for the circle
4) Change the number of hits to anything other than what they already are
5) Change any ONE of the nominals to just about anything you want and make sure the vectors are correct, fix if not.
6) Tell it OK and it will then ask if it is OK to equally space the hits, tell it yes
7) Repeat 2 through 6, except put in the actual number of hits you want to use and the correct nominals and again tell it OK to equally space the hits.

I use this for 'flat and sqaure' circles and it does a good job of putting all the hits right where they need to be. What can it hurt to try. After you do it 50 or 1000 times it gets to be quite quick and easy!

[Paul Sarrach]

You are using prehits?

[Shelley Liebman]

I will try as a cylinder, and see what happens. I did try to change the nominals and let it change the hits, but it went back to the original wrong vectors.
Paul, what do you mean, "am I using prehits", are you asking whether or not I have sample hits for the feature? If so, it wasn't an auto circle, so I do not have any sample hits.

I had to temporarily stop with the program as my mind was wondering......and I had to implement a new probe tip, so have to go through a full calibration.
It will take me at least 2 hours to try the cylinder thing, but I will let you know cause I have to figure this out one way or another.

Thanks for the quick reply guys, I'm sure I will have more questions soon!

[Wes Cisco]

compound angles

I have had quite a few compound angle holes of late, here is how I handle them:

1) I put a best fit gage pin in the hole, protruding at least 1". I work the phd 10 (adjust tip angles) until I have the AB combination that comes closest to matching the pin. Make sure you view from two sides.

2) Remove pin. ( very important! LOL)

3) Manually probe a cylinder with at least 4 hits per level at 2 levels at least .1" apart. Try to keep the levels square to the hole, but don't worry too much if they are a little out. Hit done.

4) Now using print dims for hole start location and size type in an autocylinder, but use the vectors from the manual cylinder. Create the autocylinder so it will be measured dcc.

5) Delete the manual cylinder.

6) F9 the autocylinder and correct nominal vectors to match the actuals. You may have to do this a couple of times. You may not ever get it perfect, but when you get to the point where you seem to be going back and forth, you are close enough.

It is a bit complicated. I rarely get CAD models and usally the print shows the compound angle as two single angles in two different views. I have had good results matching print angle to X.XX degress with this method. Change workplanes and angle dimension cylinder to axis for each angle. Often you have to toggle (F7/8) for the compliment of the angle pc-dmis spits out. Depending on hole size and depth you may have to limit your start stop angles for the autocylinder so your probe does not shank out. HTH

[Shelley Liebman]

Wes, I will try your suggestion, as it seems logical that it should work. And it is good to know someone else out there who has difficulty obtaining CAD files for parts, I rarely get them, and when I do there always seems to be something wrong with them at the time that I am writing the program, hense I have to teach the machine anyways!

[george frick]

on that particular hole you can just rotate your workplane 45 degrees and it will work. .7071= sine of 45 degrees.

[John Kingston]

Wes, I will try your suggestion, as it seems logical that it should work. And it is good to know someone else out there who has difficulty obtaining CAD files for parts, I rarely get them, and when I do there always seems to be something wrong with them at the time that I am writing the program, hense I have to teach the machine anyways!

cad n.

A man whose behavior is unprincipled or dishonorable.

I feel for you Shelley, above definition is the closest I get to cad at work.

[Wes Cisco]

How did it work out for you?

Yes, rotating your alignment and using 0,0,1 for a vector is easy enough when it is a single angle, or 45s but what would the vector for your cylinder be if the angle is 22.78 degrees from the y axis and 31.6 from the z? I will post praises to you often if you can tell me how to calculate a vector like that from the constituants of the compound angle. Using my method, once I have measured the autocylinder I can use 2D angle to axis in each work plane and they come out right. Please, if anyone knows a formula to calculat such compound vectors, I would love to see it. TIA

[Scott Brown]

Compound Angles

Hi Wes

I think I may be able to help with the compound angle issue. I have a formula that will help calculate the third angle of rotation so that you can square your alignment up with a hole that is on a compound angle simular to the method described above for a 2D angle. (3rd angle) 1/tan = (angle#1) cos * (angle#2)tan
3rd angle) 1/tan = 22.78 cos * 31.6 tan
3rd angle = 29.56

Step #1 Place origin at theo(xyz) center of hole
Step #2 Rotate about X 22.78
Step #3 Rotate about Y 29.56
Step #4 Measure Circle at X = 0, Y = 0, Z = 0 , I = 0, J = 0, K = 1
Step #5 Recall original alignment and dimension circle.

You may have to experiment with the angle#1 and angle#2 order. I'm just guessing without seeing the print. If you would like to check to see if you have done the proper rotation, create a generic plane in the new alignment with vectors (0,0,1). Recall the original alignment and measure the angles between the plane and the appropriate axis. (should get 22.78 and 31.6)

Good Luck
Scott

[Wes Cisco]

Mucho Gracias, Danke Schone, Thank You Very Much!

Scott,
Thank you. I just made those angles up, but they are very similar to the ones I have been dealing with, and will see again soon. I will try your formula at first opportunity.

[Shelley Liebman]

Using the cylinders worked, usually I shy away from cylinders cause they seem to cause me many headaches. But it appeared to have worked---so far.
Thanks for all the help.

[Paul Sarrach]

Yes Shelley I was refering to sample hits. Why can't you use autocircle?