Bill of Materials

[plowe68]

We are a medical device manufacturer since the 70's. With people coming and going our drawings and BOM structure has evolved into two different groups. Group 1 drawings (traditional) for example would be of a molded component that would specify the material in the notes. Group 2 drawings (new) for example would be of a molded component that would include a BOM, where the BOM would have the material as an item. That material item would be its own drawing and in that drawing we would specify the material details.
All other ancillary groups here are capable of dealing with these two groups.

Is there a direction the med device industry is going where group 1 or group 2 would be better suited?
Any focus by FDA on this subject? Or other auditing bodies that would be a red flag to them?
I didn't find a drafting requirement that specifies how BOM's are to be structured. Does one exist?

any feedback is greatly appreciated.

 

[drawoh]

plowe68,

I see two scenarios with moulded components. You order the moulded component from a vendor. They mould it and ship to you. Alternately, you order the component. Somehow, it gets moulded, then somebody does secondary work on it like machining and/or installing thread inserts. In the second case, I would expect to see a BOM on the drawing, especially if the secondary work is not done by the same vendor.

You need to manage inventory and write purchase requisitions. Which of your processes supports this better?

--
JHG

 

[looslib]

See if anything in ASME Y14.34 will help.

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

Ben Loosli

 

[MadMango]

When I worked in medical devices we had engineering specifications for the materials, and called that spec out on the general notes of the drawing. Is GMP still a requirement?

https://www.ispe.org/initiatives/regulatory-resources/gmp

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[cowski]

I work with a lot of plastic injection molded parts/drawings, though not in the medical industry, so the below may not apply to you...

We adopted the material drawing and BOM (material callout is component of part) as a best practice to facilitate changes. If a material is discontinued or we find an alternate supplier, only one drawing (the material drawing) needs to be revised. If the material is called out on each individual drawing, a material change could require a revision to dozens (or hundreds) of drawings. Additionally, doing a "where-used" on a material is much easier if it has its own part number/drawing and BOM structure (on our outdated system, at least).

 

[drawoh]

cowski,

If you change the material specification on your drawings, would that not require a revision on each and every drawing? If your material is that complicated, you could assign your own part number to it, and create a specification control. Your drawing would call up your number. The specification control would show all the acceptable materials, and would be updated without affecting the fabrication drawings.

--
JHG

 

[SwinnyGG]

If a material is discontinued or we find an alternate supplier, only one drawing (the material drawing) needs to be revised. If the material is called out on each individual drawing, a material change could require a revision to dozens (or hundreds) of drawings. Additionally, doing a "where-used" on a material is much easier if it has its own part number/drawing and BOM structure (on our outdated system, at least).

Based on your explanation, I understand why someone might want to do it this way- but to me this seems dangerous.

If a material is NLA, supplier is changed, whatever, and you edit a single document which pushes that material change across dozens or hundreds of components, how do you know that this new material is a safe replacement for all of those parts?

Unless I'm missing something.. you can't. Seems risky.

 

[drawoh]

jgKRI,

This is why I like a specification control. The spec. control should specify maximum stresses, elasticity, coefficient of friction, maximum temperature, colour, RoHS, QA and any other requirements, then list materials that are known to comply with them. A vendor can read the specification and propose a new material.

--
JHG

 

[SwinnyGG]

Drawoh,

Your proposed scheme is definitely an improvement- gives you much lower odds of shooting yourself in the foot.

I think my perception of this is colored by a few experiences where non-engineering people made material changes to parts for non-engineering reasons (cost.) and were able to do so because of a lack of engineering control over certain parts of the procurement process. So I'm extra gun-shy about materials changes to existing parts.

This still feels like a potential source of mistakes- if you're doing very intricate engineering, where specific properties matter to a level beyond 'is the Sy above this minimum' but that's probably rare. Seems like a reasonable approach.

 

[drawoh]

jgKRI,

How did the non-engineering people make changes? Did they provide verbal instructions to your vendor without telling you? Did they write an ECR so that a CAD[ ]monkey could update the drawings, again without telling you? If it matters, you must control your documentation, and your vendors must follow your documented instructions.

--
JHG

 

[3DDave]

drawoh - I got blamed when a project manager decided to send an optical assembly job to an unqualified supplier because that would save the cost of shipping components to one who was qualified. After one window fell out during thermal testing, the test guys put tape to keep the windows in place; then they took it off because the tape meant the test wasn't real. Soon after the other window, a germanium one, fell out. No need to ask the engineer. Project management has it covered. It turns out germanium is not resilient and a few grand was shattered into pieces. Since the drawing pointed exactly where the adhesive was supposed to be used and the supplier didn't apply it there, it was my fault for not being more specific. The previous supplier had managed to do the job correctly.

On another program a program manager just waived all sorts of requirements. At first we thought the suppliers must be lying but after the third different supplier says "Your program manager said it was OK," it easy to notice a pattern. He was off the program by then, but not before all the orders had been placed. It took more than a year to discover and correct all the problems he caused.

 

[cowski]

If a material is NLA, supplier is changed, whatever, and you edit a single document which pushes that material change across dozens or hundreds of components, how do you know that this new material is a safe replacement for all of those parts?

When the material has its own part number/drawing, and the plasic components have a material BOM structure, we can easily do a "where-used" to find all the parts that make use of the material. Engineering can look at the proposed material, and quickly see what parts would be affected. After some due diligence, the material can be accepted or rejected. If the material is called out on each individual drawing, a "where-used" report is very difficult to generate to see what parts will be affected (in our system, at least - hopefully your system is better).

And again, I don't work with medical devices, so this system may not be for you. But in small consumer goods where the exact material properties are not crucial, it works well.

 

[SwinnyGG]

where the exact material properties are not crucial, it works well.

This is the crux.. ultimately, if it works and you're not doing product which is life safety critical, or subject to detailed materials documentation, etc, then there's no reason to make your process any more complicated than it has to be.

How did the non-engineering people make changes? Did they provide verbal instructions to your vendor without telling you? Did they write an ECR so that a CAD monkey could update the drawings, again without telling you? If it matters, you must control your documentation, and your vendors must follow your documented instructions.

My gun-shyness is the result of past lives- in my current situation, documentation is well controlled and I have to worry about verbal agreements from other parties much less; this environment is also one where if failures happen, people can get injured or killed- so its a lot easier to get people to adhere to best practices.

In the past, I've been in environments where saving pennies was seen as a higher virtue, in certain cases, than building good product. Which resulted in a lot of shady things happening to my drawings after they were 'frozen' and released.

 

[SwinnyGG]

and the plasic components have a material BOM structure, we can easily do a "where-used" to find all the parts that make use of the material.

Can your whereused function read values that are stored inside part documents? (for example, if you're in the solidworks system, can you read custom properties)

If so, you could set the material as a custom property. Maybe you're doing this already- but if you aren't, this method combined with Drawoh's scheme for managing material properties could be very powerful.

 

[cowski]

Can your whereused function read values that are stored inside part documents? (for example, if you're in the solidworks system, can you read custom properties)

No, the current system used for BOM control knows nothing of the CAD system (unfortunately).

 

[drawoh]

drawoh - I got blamed...

I think you are agreeing with me. This is the inversion of Spiderman's theme. With great responsibility, you had better have power. If your organization will not follow your instructions, you cannot manage risk. Perhaps the replacement of germanium windows and the schedule delays are part of the cost of a nimble, fast moving organization. I hope your organization is not responsible for safety.

--
JHG