What's new

Question For The Sheet Metal Guys

Idahoan

Member
When calculating K Factor, is the "bend radius" the actual formed radius on the part, or the the tip radius of the punch?

Thanks,
Dave
 

Idahoan

Member
I'm somewhat new at doing sheet metal. Most of the time I program my parts from a flat pattern that uses outside bend deduction. I have been playing around with using K-factor just to see if I can get some correlation with the size of the flat pattern. The program that I use to flatten the part and program the NC punch doesn't always play nice with the 3D models, it would be nice to be able to flatten the part in the CAD program and bring it in to the punch program as a 2D file. This is at my day job and we a different CAD program, I thought I would ask here as there are some pretty sharp guys on this forum.

I did find some information that answered my original question that it is the radius of the part and not the punch that is important.
 
Idahoan -- The "K-Factor" was created by Lockheed during WWII to allow the untutored to generate "loosely accurate" sheetmetal parts for the era's aircraft. Tables of "K-factors: existed well into the 1960's. However using the "Neutral Axis Radius" is far simpler and more accurate. [You have to calculate it anyway to generate the "Lockheed K-Factor" which actually complicates things!]
 

Attachments

  • K Factor.JPG
    K Factor.JPG
    25.2 KB · Views: 40

JST

Alibre Super User
The way I have always done it is with this. The K-factor is related, but is somewhat of a "real information deleted" approach. This table has always worked quite well, and the formula given has worked for pieces outside of the table. The K-factor annoys me.

CCF07132015_0001.jpg
 
Should anybody be interested, the attached image shows how to calculate the value of the Neutral Axis Radius
 

Attachments

  • Neutral Axis.JPG
    Neutral Axis.JPG
    54.7 KB · Views: 36
  • Like
Reactions: MKR

JST

Alibre Super User
I have the gift...

Beans... you know the S/W SHOULD offer the choice.

And you know we know it.

It's one reason I do not use the sheetmetal part very much. It annoys me to have to do things backwards, and I am never certain it was done right.
 

leeave96

Senior Member
When calculating K Factor, is the "bend radius" the actual formed radius on the part, or the the tip radius of the punch?

Thanks,
Dave

I don't calculate K-factor (working within Alibre or Solidworks). I use a table (they are all over the internet) for various metals and input that into the CAD, make a flat pattern in the CAD drawing. Maybe it's not the most precise, but I've been getting along with it for years. Don't know if that answers your question or is even close...

Good luck!
 
I don't calculate K-factor (working within Alibre or Solidworks). I use a table (they are all over the internet) for various metals and input that into the CAD, make a flat pattern in the CAD drawing. Maybe it's not the most precise, but I've been getting along with it for years. Don't know if that answers your question or is even close...
If you look at either of my (above) "previous attachments" orr at the "attachement" here, you may learn something.
 

Attachments

  • Neutral Axis.pdf
    28.2 KB · Views: 18

JST

Alibre Super User
It seems the K-factor can be related to a "real" number.

Seems like the K-factor was invented to give shop floor folks, who might likely not be math or trig whizzes, an easy way to get things kinda close for most uses.

You can probably work in a shop for 40 years, and need nothing but that, that and a willingness to adjust a bit by trial and error when some fussy customer wants it closer. I've worked with places like that, and they got the job done.

Problem with it is that in CAD, you don't get to try it, and the dims you come up with are treated like gospel truth. So it seems that the input to the program should be a bit better than that.......

But..... It is what it am......
 
Top