HELP! Order of product design development 1

[doldsen]

My company is looking into what the "best practices" are for doing product design in NX 7.5 based on the current tools available. Historically, we did "top down" design focusing on the finished item and letting that information drive the process part geometry (such as rough castings). Since we initially went to UG/NX, many users reversed this process and began by designing the rough casting, importing and wave linking the component into the finished file, and then "machining" the finished part using sketches/features to achieve the final model of our parts. We have seen some problems with this from our New Product Development group and would like to return to the "top down" process since new tools such as Synchoronous Modeling make it much easier. However, we'd like to follow "best practices" so I'd like to hear how other companies have dealt with this dilemma. Thanks in advance for any and all assistance!

 

[looslib]

What problems have you experienced with the casting/machining sequence?
If you think about the design process, the finished part is designed first, then the method to manufacture that design is selected based on quantity, lead-time, finish, etc. If the manufacturing process is to start with a casting, now you can add material to the surface that will need to be machined and decide which ones can be left as-cast, just loosen up the tolerance on those.
I have never encountered a design process where the casting was designed and then engineering said "What can we make from this hunk of metal"?


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[doldsen]

Agreed; my argument is pretty much the same but others within our organization are viewing the process from a more manufacturing oriented perspective.

The contention they have is you must know the geometry of your casting in order to know your finished part geometry, and since it's easier to remove material from the model then it is better to import the rough casting model into the finished cast file and perform all necessary "manufacturing operations" in order to turn the rough model into the finished.

It seems backwards from the design mentality I'm familiar with and more akin to manufacturing engineering processes, but I thought I'd get feedback from other users at other companies to make sure I'm not missing something.

 

[looslib]

While working in a conceptual world, the casting/finished part are envisioned together. But from a design standpoint, you have to know what your finished product is before you can do the casting. I don't see how they can say "you must know the geometry of your casting in order to know your finished part geometry". From the technical CAD point, it is easier to remove material than add, but now you have done your finished design on paper before starting the casting CAD file, or you are doing 2 unrelated files and doing more work than is needed to bring them together.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[doldsen]

Given the ease with which you can add material or modify geometry using Synchronous Modeling in NX 7.5 (which is the version we are running), moving from a finished part back into a rough casting seems to be a very easy task. Not sure if anyone else has experience in do this type of operation yet, but I'd be glad to listen to any and all feedback on the subject.

 

[jerry1423]

I agree with the last statement regarding "Syncronous modeling".
Previously I would have rather machined the casting on the CAD model, but now I find it just as easy to add material (with some exceptions) - it just makes more sense to come up with the finshed model (because that is what you ultimately want) and create the casting (add material, remove holes) from there.

 

[capnhook]

You WILL be up-scaling the model for material shrinkage, won't you?

jerry and looslib are correct. I have a lot of casting tooling experience, and have always started with the finished design, added material, removed holes, and then applied scale.

Proud Member of the Reality-Based Community..

 

[doldsen]

Yes, and traditionally they have managed this through the Scale function in the Transform menu. I've been trying to demonstrate the capabilities of using Offset Face in Synchronous Modeling to "reverse shrink" a casting, but so far it's been tough going for some reason.

I thank everyone for their feedback as this has been exactly the type of "ammunition" I need to argue design standardization to upper management based on top-down design processes.

 

[ewh]

I'm curious, how is offset face preferable to scale for cte?

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[JohnRBaker]

If you're attempting to compensate for the 'shrinkage' which occurs when casting or hot-forging a part, then using SCALE is the only approach which will come close to duplicating this. And note that the Scale Body function supports Uniform, Axisymmetric and General (X,Y,Z) scaling so that you can even account for the effects of shape such as large length:width ratios.

Offset is fine to compensate for material removed while machining, but it's not representative of how an object changes size as it cools.

John R. Baker, P.E.
Product 'Evangelist'
Product Design Solutions
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.com/museum/

To an Engineer, the glass is twice as big as it needs to be.