Engineering Design and Sheetmetal Programming ?

Just out of curiosity, how many of you SWX designers do your own sheetmetal
programming? I work for a large firm with 10 Engineers and 4 of us use SWX
for design. What they invision is having us designers doing our own
programming for the Amada for every peice of sheetmetal we create for our
different products. At this time, we have one person who does not use SWX
and is not a designer. He has been doing all our sheetmetal programming
using Fabriwin for many years and is fairly fast at it. But he cannot keep
up with the pace of work. At the same time, us as designers can barely keep
up with our design duties due to the new product lines and demands for new
parts on a daily basis. I personally think they need to hire a hot-shot
programmer (younger than the one we have) to take on some of the programming
responsibilities for the sheetmetal. Someone with some spunk who would like
to work his way up in a good company. The company on the other hand thinks
the parts will go thru alot quicker if we also program them. Am I crazy for
thinking they are dead wrong.?? I think it will only slow us down on the
design end of product and things will be released to production slower
rather than faster. My feelings are a person has to excel in a certain area
of the chain. What I mean is I am fairly proficient with SWX and when not
burdened with other things, can get thru a design rather quickly. Same thing
with the programmer. He does not have to worry about designing the parts and
therefore, he can excel in his progamming chores. What is everybodies take
on this?? Just as a side note, I have programmed up to 5 -axis machining
centers and lathes with live tooling for over 25 years, so I am no slouch
with G-code or any program related to creating it.I just think they are in
the wrong mode of thinking when heading in this direction. Am I wrong. Dont
be shy. Answer honestly because It wouldnt be the first time I have been off
base. I have been in maunfacturing for approx. 34 years, so I have been
around a few places and seen good ideas go sour a few times. Thanks for the
input....Jake Barron

 

I would have to agree with you. Think about the details of the post
processor, tooling selection, nesting etc. I don't know how much you make
one/multiples of a particular part per sheet, but if you nest various parts
on a sheet, that certainly swings the pendulum to the programmer side.

OTH, making sure your SW flat pattern is correct is probably the quickest
way to help the process. Develop your k-factor, etc. so that you get good
results.

WT

 

I do both. But I do most of the programming of our sheetmetal parts, we
have a turret and a plasma. First of all won't the cost of 10 fabriwin
liscenses alone scare them off. I don't know how complex your parts are but
using fabriwin on SW parts is easy and takes only a minute or 2 to program
for a linear machine. Though for the turret it is quite a bit more
cumbersome and may take 30min or more to do a single part. We also have a
seat of InteliNest for the Plasma machine, this is quite the time saver. It
also allows us alot of versatility since we are able to fill a days worth of
the nesting in 1/2 hour, though our shop leaders do the daily nesting so it
isn't an engineering task. If your machine is a plasma or a laser I
wouldn't worry about the time it takes to program a part in fabriwin as long
as you don't have to nest it too it may take you an extra 10 minutes if that
to do your own parts. If you have a turret I would be worried because it
could add quite a bit of time to your tasks. From AutoCad it is a different
story because you have to clean up each drawing. I have found that for the
most part an AutoCAD part if it is drawn to scale and hasn't had any "Cheap"
editing done to it, it imports quite well and quickly maybe 3 minutes per
part. But if there has been cheap editing it could take a long time because
you have to make the geometry correct. Some things are easier to correct in
Fabriwin and others are much easier from AutoCAD. There are some quirks to
Fabriwin that you will have to work through but once you know what to look
out for you can work around it pretty easily. (Watch out for any part that
combines "Old" Sheetmetal techniques with "New" Sheetmetal techniques).

Anyway in short if you have a linear cut machine it will be a brease from
AutoCad and SolidWorks, but if you have a turret it will suck.

Corey

 

From someone who does it all i woulsd say that you really do need a
dedicated person to program. I use SMP from Merry Mech to program an
Amada Turret punch. Every part that i design i have to tool up myself.
This can take quite a while and in turn it slows down the design work.
On the other hand it can be useful to be doing both as you have
complete control over the part being designed. But in short have
someone else to write the progams would be better

 

I do some design from the ground up, but primarily use SWX as a
QA/Pre-production tool to prepare documentation including flat
patterns. I support two turrets, a laser and 5 press brakes for two
shifts. Jobs are coming in from 3 quoters.

As designers, you should be able to provide the programmer all the
documentation, including an accurate flat pattern without too much
difficulty. From there, he should be able to concentrate on CAM and
lay-out. While I use MetaCAM, Fabriwin would be my 2nd choice and is
entirely capable. I currently handle all documentation in addition to
programming, so I'm wondering what kind of load you are placing on the
programmer. 10-20 new parts a day?

In any case, it's more cost effective to add a junior programmer than
to teach CAM to engineers. The tenth time tooling gets side loaded and
ends up ruining punches and material, they may realize this. It's not
rocket science, but in the situation you describe, having engineers
program the punch makes as much sense as having you puch the parts as
well. In fact, many people believe that it takes an operator to be a
decent programmer. Knowing the machine, its quirks and capabilities is
half the job.

Good luck

 

OTOH (and even though I agree), what seems on the surface to be for the
good for the company may not be quite as good for the employees . . .
and MAY not even be as good for the company as it seems it might be.
Engineer who can program CAD can do better sheetmetal design. If the
engineers are tasked as planned, the slowdown in new design will be
noticed by management, and some type of adjustment will be made, whether
for the good or the company or not for the good of the company. Unless
it results in a layoff, it may still end up being better for the
employees as they end up with enhanced abilities. Perhaps (and I'm only
saying this for sake of playing Devil's Advocate) the engineers would do
better not to look a gift horse in the mouth.

Mark 'Sporky' Stapleton
Watermark Design, LLC
www.h2omarkdesign.com

 

Quite well put, I might say. Back in 1980, our engineering department
purchased a laser/turret punch for making prototype parts. I was not in the
parts making end at the time, but kept pestering the boss to let me learn
it. I figured that a new toy was an opportunity that should not be allowed
to pass. So, I read the book on my own time, got some time under my belt,
eventually taught the others a thing or three, and ended up running that
department.

So, while I still think a dedicated programmer will be more efficient
overall, having the designers understand what goes into producing the part
they design, most certainly makes for better design. And to personally not
seize an opportunity to learn something new on company time would be
foolish.

WT

 

Im not shure, but what I have seen, is that the Amada add-in for SW, is that
it allmost does all unfolding and programming with just a few klik on the
mouse.

Ofcorse this has been seen at EuroBlech (Exhibition i Germany), were every
thing works first time.

Sorry my english

Jørgen

 

We use SW2005 and Sigmanest to laser cut all of our sheetmetal parts here at
Lakeside Manufacturing. We have three designers here and all three of us
here create our own laser files for Sigmanest, but we have a dedicated guy
doing the nesting out at the laser. It seems to work real well for us. It
allows us, the designers, to verify our blanks before they get nested.
Sigmanest also has a Solidworks module that allows us to import sheetmetal
parts right from Solidworks. It also allows you to open a full assembly, it
then traverses the assembly looking for any sheetmetal files and lets you
set them up one after another, without having to import each file
seperately.

Richard

 

Jake -

I let the designers design and the programmers program. It makes no
sense to have it any other way.

The designer should know enough about the process to be attentive to
"downline" problems, but should not have to worry about programming a
dwell into a corner, when to toggle the assist gas or tweak the wattage
for a given material, activate trap door or whatever.

The programmer should not worry if the flat is correct, all the
features are present, etc.

I would suggest that they get a motivated person to play the role of
designer programmer and let them take up the "balance" of the work as
needed. I worked in a group of designers & programmers and each of us
had a specialty with an overlap in another area. Some more that others
based on ability. I believe that letting a programmer get into part
design is a great thing and vice versa, but it will not make the
process more productive to have everyone following the same process "on
their own".

Taken to a crazy level, we need to have them purchase and source their
own flat material, receive it, design a part, program it, run it, brake
form it, paint it, package it, deliver it to the customer and then
collect the money and split their portion with the company . . .

Some companies run like this, but generally have 2 to 3 workers.
Economies of scale need to stay economical.

Later,

SMA

 

Jake -

I let the designers design and the programmers program. It makes no
sense to have it any other way.

The designer should know enough about the process to be attentive to
"downline" problems, but should not have to worry about programming a
dwell into a corner, when to toggle the assist gas or tweak the wattage
for a given material, activate trap door or whatever.

The programmer should not worry if the flat is correct, all the
features are present, etc.

I would suggest that they get a motivated person to play the role of
designer/programmer and let them take up the "balance" of the work as
needed. I worked in a group of designers & programmers and each of us
had a specialty with an overlap in another area. Some more than others
based on ability (I did Tooling desing, prototype support & VMC
programming in that order - another guy did Laser Turret Programming,
Prototype support & tooling, etc.). I believe that letting a
programmer get into part design is a great thing and vice versa, but it
will not make the process more producive to have everyone following the
same process "on their own".

Taken to a crazy level, we need to have them purchase and source their
own flat material, recieve it, design a part, programm it, run it,
brake form it, paint it, package it, deliver it to the customer and
then collect the money and split their portion with the company . . .

Some companies run like this, but generally have 2 to 3 workers.
Economies of scale need to stay economical.

Later,

SMA