Scaling a part in Y question

[jeff]

We often do shafts out of 8620 that are 24-40" long or longer, and we usually have to account for .001"/inch "stretching" on the length when carb and hardened.

Is there a quick way to use the Scale feature to make the part longer by this amount per inch?

I'm having a brain fart at the moment.

 

I know it's not normal to make a part out of tolerance by .03",.but it works lol

 

EDIT: topic should read scaling in X, not in Y

[jeff]

I think I got it,

Scale in XYZ

X 1.001

Y 1.

Z 1.

[JoshC]

Yea there may be some math, i believe its A/B=C, where A= Size we want it to be B= size it currently is and C=factor ratio.

So with the math problem listed above you can just replace the variables;

​

for example we have a part that is 5" long in the X and 5" tall in the Y direction and lets say we want the new x length to be 5.01.

the scale function would look like this, 5.01/5=1.002 --> so in this case our scale factor is 1.002.

​

Edit: or to be more specific to your question.

our parts 24" long and we want it to be 24.010" long. the scale factor is 24.010/24=    1.000416666666667

[themachinist]

EDIT: D'oh! Comment below is about G51

 

What kind of control is it? There are a couple of gotchas.

• For instance, on Haas controls you cannot scale by a vector. You can only scale the entire pattern (i.e. all axes).
• Some (all?) Fanuc controls will use IJK for the scaling factor of each axis (XYZ respectively) or P to scale all axes simultaneously
• On Fanuc controls there may be a parameter you need to enable to allow scaling of an individual axis.

There may be more depending on the work and remember that you cannot enable scaling while cutter compensation is active.

[civiceg]

Why not just move your wire frame/modify your model. I would always add a note on the level added XX to this feature. 

 

I hate making an educated guesses with things like weld shrinkage and elaborate plantings.

[guitar]

You answered your own question Jeff. I used to scale my model the same way when making molds.

[jeff]

What kind of control is it? There are a couple of gotchas.

• For instance, on Haas controls you cannot scale by a vector. You can only scale the entire pattern (i.e. all axes).
• Some (all?) Fanuc controls will use IJK for the scaling factor of each axis (XYZ respectively) or P to scale all axes simultaneously
• On Fanuc controls there may be a parameter you need to enable to allow scaling of an individual axis.

There may be more depending on the work and remember that you cannot enable scaling while cutter compensation is active.

I'm doing it in Mcam X5  

 

We have a few different lathe controls, Okuma P200L, 0TB, MSD501, P300L, some crappy old haas, and an older Yasnac control. 

[Mjölnir]

You answered your own question Jeff. I used to scale my model the same way when making molds.

 

caveat. when scaling values are not the same along each axis, arcs become conical. is not an issue for surfacing molds. is a significant issue for making features which need to be round, round.

[jeff]

caveat. when scaling values are not the same along each axis, arcs become conical. is not an issue for surfacing molds. is a significant issue for making features which need to be round, round.

I didn't think about that, I would only be scaling a 2D  wireframe of shafts with a bunch of journals,undercuts, and the typical radii and chamfers.

[JoshC]

caveat. when scaling values are not the same along each axis, arcs become conical. is not an issue for surfacing molds. is a significant issue for making features which need to be round, round.

Yes, but in theory it would only be non-round until it shrinks, if the model shrinks as it should they will become round again after Heat Treat. So to correctly manufacture a part that will shrink more in a length than in a width direction the part would be nonround while being machined and in theory will shrink into the correct shape.

[crazy^millman]

Yes, but in theory it would only be non-round until it shrinks, if the model shrinks as it should they will become round again after Heat Treat. So to correctly manufacture a part that will shrink more in a length than in a width direction the part would be nonround while being machined and in theory will shrink into the correct shape.

 

Or you make the critical features undersized and finish them to size after heat treat to know it is correct and work with reality verse theory,

[crazy^millman]

Yea there may be some math, i believe its A/B=C, where A= Size we want it to be B= size it currently is and C=factor ratio.

So with the math problem listed above you can just replace the variables;

​

for example we have a part that is 5" long in the X and 5" tall in the Y direction and lets say we want the new x length to be 5.01.

the scale function would look like this, 5.01/5=1.002 --> so in this case our scale factor is 1.002.

​

Edit: or to be more specific to your question.

our parts 24" long and we want it to be 24.010" long. the scale factor is 24.010/24=    1.000416666666667

 

Had a topic about this very thing about 3 years ago and about 8 years ago. We pretty much answered it the same way. Good to see people using math the same today as it was back then. No if we can get others to see yesterday's math is the same as today's then things might get easier on some other fronts.

[JoshC]

Or you make the critical features undersized and finish them to size after heat treat to know it is correct and work with reality verse theory,

 

But everything works in theory...

[crazy^millman]

But everything works in theory...

 

Yeah I had some people making theories a about 3 month project only taking 2 weeks. Guess how long it really took?

[mikenaturalice]

Yeah I had some people making theories a about 3 month project only taking 2 weeks. Guess how long it really took?

Man o Man! I know that feeling. Had a team of consultants estimate a project a t 18-24 months, so of course the powers that be decided that was 'overkill' and went with 9 months. 

[Mjölnir]

Man o Man! I know that feeling. Had a team of consultants estimate a project a t 18-24 months, so of course the powers that be decided that was 'overkill' and went with 9 months. 

 

 

I love it when tools (injection molds) with 3000 hour lead times sit in engineering for 6 months and are finally released to the shop floor to build 6-8 weeks before their due dates.

[Thee Rickster ™]

I love it when tools (injection molds) with 3000 hour lead times sit in engineering for 6 months and are finally released to the shop floor to build 6-8 weeks before their due dates.

Hence the stress of a mold maker.