Sheet metal: length on drawing different than model

[bwilliams-4]

Hello all,

I have a question, to which I cannot seem to find an answer.  I am hoping y'all will be able to help me.

In my work, we use plate steel and bend it.  Pretty simple.  I use the Sheet Metal application (Creo 3.0, M090), so that I can flat pattern the parts and be able to tell the manufacturing people what the overall length of the part they will laser out needs to be.

I sketch the part in its bent state, and I have been taught/told as best practice (or so I believe).  Make some cuts, and I have what I want.  I sketch a curve that is half the thickness, and then created a parameter, as well as an analysis feature, and tie said parameter to the measurement.  This was, I know what my length SHOULD be.  (see my first slide in my attachment)   In the first slide, its just a 3D view, and the sketched curve is there to be seen.  I have the curve projected to the same height as the points at the ends.

When I put the part on a drawing, I noticed that the dimensions I get by doing the math do not add up.  And I have measured what the length should be several different ways.  One was via AutoCAD, one with MathCAD.  I can get those values to match up.  And the measured length from the sketched curve I put in as a feature yields results that are close (but off by 0.009 inch)  (and yes, I get it, that's close enough to call it good enough).  (see my second slide)

So my questions are:

Why is my flat pattern, point to point measurement, so far off (it's almost 3/16ths different)?

Is this related to the K factor and/or Y factor?  If so, how do I manipulate those values, because I have no clue.

Why is the sketched curve off from what I calculate?

Is there a way to get the flat pattern to be true to the length I calculate?

Thank you in advance to anyone who can help me shed some light on this.

[dschenken]

The K or Y factor affect this.

In bending it is not possible to compress the metal on the inside of the bend so it has to squeeze sideways which is hard to do. On the outside of the bend it stretches and becomes thinner. Were the metal ideally elastic (which cannot be the case for plastic deformation to work) there would be equal amounts of compression and extension, but it is not, so the effect is unbalanced.

Rather than being the length of the mid plane, the neutral/unaffected length is pushed closer to the inside of the bend in various amounts, depending on the method of fabrication and the extent of the bend. In stretch forming is is even possible to shift the neutral length within the inside radius.

A start value for Y factor is .50 but it really depends on the tooling and the material. See SheetMetal.Me – K-Factor Same thing with springback, which is also not included in these calculations.

Here's a good article SheetMetal.Me – Y-Factor

[StephenW]

In addition, you may want to refer to machinery's handbook, they have a section on bending. If you know how the part is manufactured (air bending, bottoming die, etc) and you know the material, you can get some really accurate results based on formulas.  If you are sending parts to arbitrary vendors to be manufactured, it may not be a good idea to send flat patterns since you may not know the process they are going to use. If you must send flat patterns, make sure there is a disclaimer about the accuracy of the flat and it is the vendors responsibility to verify all lengths are correct based on their processes.

I have used the centerline method way back. It can be close on thin sheet. It get's wildly inaccurate on thicker plate.

[BenLoosli]

Never send a vendor the flat pattern!!! You are not buying that intermediate shape, you are buying the finished bent part. IF you have a good working relationship with a vendor AND you understand their process, then send it as a reference file only. I have been caught with rejected parts in the past where the vendor used our flat pattern and then the final shape was off.

Search online for a paper called BendWorks, The fine art of SheetMetal Bending by Olaf Diegel. It has a good explaination of K and Y factors.

Creo can develop the flat pattern for you without constructing the mid-plane sketches.