Design & Manufacturing 7

[MadMango]

I would have titled this Design for Manufacturing, but I didn't what to bias people right off the bat.

I'm just looking to get a feel for what others are doing. This question deals with sheet metal design and in-house punch press fabrication, but could apply to anything designed where your resources are limited to certain manufacturing processes.

I'm wondering if you have created pallet parts or library features (or other) of your standard punches and dies for in-house fabrication, or design from a list of available sizes, or if you design what you need and let your manufacturing department worry about meeting your specifiactions?

In this day and age of competitve markets and constant drive to produce the highest quality item at the lowest possible manufacturing cost, it dawned on me that I should be designing with manufacturing (as well as assembly, maintenance, asthetics, durability, quality, etc) considerations clearly at the forefront on my mind. I'm just curious to see how others are addressing this.

[green]"But what... is it good for?"[/green]
Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[J-C]

I mostly always designed sheetmetal parts per what manufacturing could make. Which means using standard round (holes) and obround (slots) sizes per a list. You could set this up as a library feature now that it's supports configurations, in fact I provided to Solidworks the slot size config list that is currently in 2005.

Tehre are cases where need to push the envelope of what was currently manufacturable. We weighed those on a case by case basis. Usually meant buying more tooling or custom tooling.

Jason Capriotti
Smith & Nephew, Inc.

 

[davidinindy]

Not sheet metal, but I always design for the equipment we (or our tool shop)have.
No sense in making them go buy new stuff for each project, or try to figure out how to make something with what they have.

 

[ctopher]

Only in some special cases we use library features, etc. Mostly we leave it up to mfg. I let them do whatever is needed to do the job ... is easier and cheaper for all of us.

 

[alexsasdad]

What we have always done in the places that I worked for is we get a CURRENT tooling list from manufacturing. Then during the design process we try to use features that we have tooling for. If we don't have tooling for it we ask, "Does this feature really need to be unique or can I use a standard one?". Most of the time yoou can do it with a standard tool, but in some cases it has to be something new.

 

[CorBlimeyLimey]

I design with what is best for the product, not the machine or fabrication shop. If you don't introduce new features, techniques or processes into the product, the shops will never move ahead. Of course new features, processes, etc, should only be used if they are truly beneficial to the product ... not just for the sake of change.
Sometimes though, if time or cost does not permit acquiring the necessary tools or machines, a compromise needs to be found.

 

[MadMango]

A little more background on this topic. Our shop has 2 Amada punch presses and a 3kw laser. During the product development stage most of our parts are cut on the laser. It's faster and doesn't force a design compromise.

In Limited Production, due to production priorities and capacity in the shop, some parts get moved to the Amadas. This sometimes modifing the parts so they conform to the exisiting tooling.

It seems to me that designs (sheet metal in our case) should be approached in a manner to allow the final parts to be made on either machine, especially if one or the other goes down to failure/maintenance.

So it seems that most of you haven't gone much past just referencing a hard copy of your shops tool list, except for Gildashard. Sorry Jason, but I haven't been able to use SW05 in any major capacity yet to see your contributions.

[green]"But what... is it good for?"[/green]
Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[drawoh]

MadMango,

I am very interested in DFMA, although there is no company policy on it here.

I have bought a copy of the Doothroyd, Dewhurst, Knight textbook on it. There is a lot of stuff in there that is helpful to know.

I think the real issue is that you talk to your fabricators and assemblers and make yourself aware of what they can and cannot do. Most bad sheet metal drawings I have seen have +/-.005" tolerances all over the place. Sheet metals shops have told me they can bend to +/-.015".

I recommend doing complete, quality documentation. Production needs this anyway. When you break your assembly drawing down into clear assembly steps, or you write out assembly instructions, you pick out design flaws. A designer should never write instructions on how to winch an engine out of a car to change the spark plugs.

JHG

 

[Theophilus]

Most of my parts are plastic and intended for mass production. However, when we get prototypes (such as SLA or SLS) or even soft-tooled urethane parts, I alter the designs somewhat for those processes. I design for the intended process while maintaining the integrity of the design intent--and of course a great product.

Failure to design with the production process in mind will always compromise the quality of the product or drive up the cost when someone else downstream has to handle it for you.

I try to create parts that speak for themselves. Parting lines are implied by draft and other features so there can be no other way of molding a part except by what the features imply and require. This eliminates a tremendous amount of hassle when moving to production and I believe it is the obligation of every engineer/designer to do the same. Products simply cannot be produced without all these items being defined by someone.

Jeff Mowry
Industrial Designhaus, LLC

http://www.industrialdesignhaus.com

 

[ctopher]

>.005 tol is sort of a default tol. Engineers/designers/drafters that use it all over a dwg usually means they don't know how to tolerance a part.
It is a good idea to have some reference of machine tools to design by and tolerance the part accordingly.

 

[Dennyd]

Engineers that design without really understanding the processess that they are inherently specifying are nothing more than cartoonists. Well, 3D cartoonists. If you are designing a molded part you have to understand all about gates, parting lines, draft, pulls, shrink, porosity, etc., etc. The same thing with sheetmetal (typically design with sharp corners for a turret press vs. lasers prefer rounded corners). Designing within the capabilities of the process is required. If the process is not capable of what you want then you have to find and design for a different process.

It is expeditious and wise to design to use as much existing and standard tooling as possible. However, if your design requires something different then it has to be acquired (tooling $$).

I've seen a few brand new engineers that can drive the snot out of a CAD system - great! But if the results aren't manufacturable then it is wasted effort. DFMA, baby! Design for manufacturability and assembly is an attitude and it can be learned. It should also be supplemented with an appreciable amount of hands-on experience. You learn a lot more about machining if you have to cut chips yourself than by watching a video. The extra advantage of this is gaining insights into what the processes are capable of. What are the things that can be done, but a lot of people don't know to specify. It makes the job just a whole lot more fun too!!!

- - -Dennyd

 

[MElam]

What we did so that our sheet metal matches the tooling is this:

1. A sheetmetal craftsman (who does know SolidWorks) bent samples of each material with each tool to reverse engineer the radiuses and k-factors. I placed all of this data into a spreadsheet, along with minimum flange lengths.

2. I then made part templates for each material by placing a simple base-flange and setting radius, thickness, k-factor and relief settings. Then delete the base-flange leaving just the sheetmetal feature and save as a template. I only did this for the most standard tooling, so if the designer needs a different bend radius they must consult the spreadsheet to see what is available with existing tooling.

The templates only work with the base-flange feature. Insert bends does not pick up on my default values.

As far as changing machines, the only quick solution that would work is if you could find settings that work for both. Otherwise, maybe a macro to adjust settings on the fly?

 

[aamoroso]

Design based on what you can make, if you cannot make it you do not have a design, you have a dream. Tooling for punching can get expensive and to expect them to buy the emboss or louver you choose is not realistic from an economic standpoint. Create a set of library features and pallete features that you can use and stick with them. Any designs for outside manufacturing, you should expect to be flexible on embossing tools or expect to pay tooling costs which will raise the cost of your product.

 

[macPT]

Again, there's not one solution for this problem and best fitteded solution is always a consequence of your product, your technology, market,...

Here's what I have implemented in my company.

1- We have machines and tools for what we need to know their capabilities. The first aproach to design is to profit from having these tools (considered as standard tools). We know how to design according to these capabilities; there's not the need to by new machines/tools.

2- Even so we try not be to much attached to the actual machines and tolling available. We always give room to some innovation, as the maintenance of the standard processes and tools can result in a solution taht is not the most economic. Even during production (that is after the design process have long ended) we always keep our eyes open for an opportunity to change tools/processes (and revise the design accordingly) to get a cheaper product.

3- We involve the manufacturing departement in the design process so they can be advised of any production change (or even suggest this change). They will give cooperation with some tasks like estimate production costs, layout modifications, planning changes...

4- In fact we involve several departements in the design process as there are many things to look at (new components/materials to purchase, product transportation, new tools for test & mesurement, ...)

5- We keep standard features/palletes on the server, and SW is pointing there so everyone can use the.

Rrgards

 

[MadMango]

[blue]macPT[/blue], that is exactly how we are running.

The problems are "created" when Manufacturing decides to move part A from one process (laser) to another (punch press) due to capacity issues in the shop. I won't discuss what happens when they decide to move products from manufactured to purchased, or vice versa....

Another problem arises from this as well, and that being in the past we standardized on imperical hardware. Now the design mandate is to use metric hardware. Very few of the existing tooling are the proper sizes for the new metric-hardware populated products (hence heavy use of the laser).

I also don't want to give a false representation of this being a "major" problem. It's jut something that has been overlooked in the past. I want to start bringing DFM more into the light.

[green]"But what... is it good for?"[/green]
Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[macPT]

MadMango

We have the same problem. Sometimes, for several reasons, the manufacturing decides to change to an alternative manufacturing process.

We are not necessarely notified of this change. We discover this in cases where the quality departement reports deviations from drawings sepcs.

Then we will try to "negotiate" if this change will last forever (in this case the design will be revised) or not (in this case there's a report stating the conditions of acceptance)

Regards

 

[2ManyHats]

It's interesting that the majority of contributions to this topic appear to be from design engineers who have in-house machine shops.

The idea of having a definative list of available tooling is novel to myself, and I'm sure, many others. I design sheet metal parts based on what tooling I have seen that our contractors have, tooling that they can modify or purchase readily.

At times we will bring onboard a new contractor and often this will mean modifying the design to allow for the closest tooling available.

Plant lists that our contractors make available never include details of punch/form tools etc.

Often a design isn't finalised until I take a concept drawing to a meeting with the contractor to ensure that they have the tooling capacity to fabricate what I have designed.

As the sheet metal fabrication is contracted at a fixed per-part cost my design priorities are 1.Fulfilling functional requirements 2.Ease of assembly of the finished product 3.Per-part cost of the outsourced fabrication.

Generally the easier the design is to fabricate the lower part cost we will see, but if it's a choice between ease of assembly of the final product or ease of fabrication of the sheet metal part then I'd design for the final product assembly since this is the costing which is variable and since sales won't live with variable retail prices this would mean we have variable margins....

...are we designers or compromisers?

 

[TheTick]

...are we designers or compromisers?

Design is always compromise.
[ul][li]Strength vs. weight[/li]
[li]Style vs. common parts[/li]
[li]Cost vs. quality vs. speed[/li]
[/ul]
to name but a few...

Sometimes manufacturing must heavily influence design. Other times, appearance overrules all.

One compromise I never make: Never let your design tools' capabilities (CAD software, etc.) be the factor that forces you to simplify your design.

"Customer satisfaction, while theoretically possible, is neither guaranteed nor statistically likely.--E.L. Kersten

http://www.EsoxRepublic.com

 

[ko99]

The Tick beat me to it: To Design is to Compromise.

Let's say you like a 0.300 boss for some hard-to-define reason but a 0.320 does the job and your supplier has the tooling... it's a no-brainer.

If a 0.320 boss causes a design problem, you brainstorm what it would take to solve the problem. Then make a decision: design tweak vs $xx for tooling. The answer varies with every case.

Innovation is all about what's best for the product, not the designer's ego.

ko (

http://www.ecooling.biz)

 

[macPT]

TheTick

I agree 100%. If we do not compromise then we are artists, not designers. But then, we would have not much we could manufacture or sell (how can we make those clocks working Mr. Dali?).

Regards