Significant Standard and Part Numbers - Provisions of Alternates by Under-Definition 2

[SAITAETGrad]

Hi CM experts. I tried to search out a solution to my problem but came up dry.

For standard and commercial parts with significant part numbers, is it ever reasonable to not fully define the part number to imply alternates?

One example would be discrete fasteners which are available with significant part number lengths. In the factory setting, assemblers will size the length on installation and choose from on-hand consumable supplies. For field alterations, the parts are supplied at an oversized length to be cut down by the installer. Would it be fine to not simply exclude the length portion of the suffix?

Another example would be second sourced electronic parts where the last digit is a tolerance code. Would it be reasonable to exclude the tolerance code to allow the purchaser to select from available supply?

To be clear, this would be directly on an assembly BOM and not contained within a source or specification control drawing.

One response may be 'it depends on your standards'. If so, what advice would you provide on structuring this standard so that it is clear for all? How do you avoid drowning in technical debt when past practices are no longer accepted based on the way the wind has blown in the quality department?

There are other ways this could be attempted such as a wild card or a note. What is the best way?

Thank you for your help. I'm sorry if my question is trivial or repetitive.

 

[MikeHalloran]

For standard and commercial parts with significant part numbers, is it ever reasonable to not fully define the part number to imply alternates?

NO.

One example would be discrete fasteners which are available with significant part number lengths. In the factory setting, assemblers will size the length on installation and choose from on-hand consumable supplies. For field alterations, the parts are supplied at an oversized length to be cut down by the installer. Would it be fine to not simply exclude the length portion of the suffix?

I come from the fringes of aerospace and medicine, where the actual delivered product had damn well better look exactly like the design documents say it looks.

Another example would be second sourced electronic parts where the last digit is a tolerance code. Would it be reasonable to exclude the tolerance code to allow the purchaser to select from available supply?

In my world, many/most circuits will not perform to the design intent with component values outside of the specified tolerance band. ... which is why the tolerance is there in the first place. Allowing an assembler to select parts from a pile of junk, or allowing a Supply Chain Manager to buy the junk in the first place because it's cheaper, gets expensive real fast.

There are other ways this could be attempted such as a wild card or a note. What is the best way?

The design documents should specify what is actually required.
Every operation downstream must respect that, or the design authority will be getting stupid phone calls for years, and won't know why right away. When they figure it out, you will get a phone call you don't want.



Mike Halloran
Pembroke Pines, FL, USA

 

[3DDave]

The purpose of an engineering drawing package is to narrow the choices to those which are likely to work. Abdicating that responsibility is not a worthwhile approach and pushing the responsibility for requirements management to another organization sidles up to perpetrating fraud. That's why there are document types that allow more flexibility. If there was some other good solution that would have been standardized and put into common practice as it's a frequent need.

 

[swertel]

This is a constant problem in industry where program management or other non-technical disciplines try to save cost by assuming that downstream functions can more cheaply choose a standard. There is one word that I can apply to your situation, traceability. (Because changing the culture of the company to get them to recognize that up front engineering is the least expensive point to define requirements is often too difficult.)

If you can ensure complete traceability by allowing someone else to create as-built records, then by all means allow engineering to not provide complete definition. Frankly, engineering should be creating as-designed records, not as-built. None the less, if there is a requirement that could affect safety, functionality, performance, or similar, then that requirement should not be left for downstream processes.

Alternate Parts:
If you have an ERP, MRP, MES, or similar system that can define suitable alternates and substitutes, then you don't need an engineer print to do so. But, who defines a suitable alternate? I recommend Vendor Item or Source Control drawings (ASME Y14.24).

Fastener Lengths:
Does the engineer seriously not know how long the fastener must be that you have to leave it up to the field mechanic or assemblers? That's scary. But otherwise, I recommend a Modification Drawing (ASME Y14.24). Within this drawing, you can define the final size required and allow any length fastener to be modified to meet the end item length.

Electronic Parts:
Research Selected Item Drawing in ASME Y14.24. This allows purchasing to buy a range of electronic components in bulk yet use a narrowed selection based on requirements to fulfill the design requirement. An example of this are LEDs. Most manufactures make pretty generic LEDs. Yet, in an automotive dashboard, they have to have a very close range of brightness. It doesn't matter what the brightness is, as long as one diode isn't really bright or really dull. The engineering BOM will define that any LED within a certain brightness range is acceptable as long as all LEDs used on a single dashboard are within that range.

--Scott
www.aerornd.com

 

[SAITAETGrad]

Dave, sorry - in both cases there would be no abdication of requirements management. The few possible alternates all work to meet the design requirements.

What are the document types that allow more flexibility and which work best?

 

[3DDave]

You don't know that at some future time a new, non-usable part number will be added to a manufacturer's list.

Since your question was in the abstract, there could eventually be thousands of parts or tens of thousands of available parts.

Maybe add a note: AS AVAILABLE ON PAGE XYZ OF ISSUE QRS OF SUPPLIER-X CATALOG to actually limit the part selection to the past and current ones, but then you need to keep a copy of their catalog because they may not. And yes, I have found companies that dropped a product so hard they pitched all their docs; they were done with that business segment and could not supply answers about materials they sold.

Look - it's your business. You asked what people thought about it and if an answer doesn't please you feel free to not respond.

 

[SAITAETGrad]

Dave, I'm not looking to pull some stunt or start a flame war - honest. I'm sorry if I've offended.

I've been given a rationale by another group of specialists in our department for their practices. It's not a small matter to resolve. I've presented the rationale as I understand it and respectfully asked for a sanity check and for tips on what could work better.

Yes, I agree standards can change but the few examples I've seen so far are for stable, well established part standards and legacy second source parts with only a few possible alternates that have so far stood the test of time. What they've done is unfortunate given that a selected item or a source control document with a little more detail would have prevented the issues we face today.

I see now that how I originally presented the problem was poor. Again, I apologize.

Scott, thank you for the helpful and thoughtful post. Greatly appreciated.

It's not that the engineer is without the precise thickness. At least in older OEM aircraft repair and production drawings which I'm more familiar with in many but not exclusively all cases, diameters are specified but lengths and counts are not. Length determination is abstracted to the production or maintenance riveting process. All rivet locations are defined in the FOD but the quantity is A/R in the BOM.

 

[truckandbus]

In my past experience we ran some bills of material without hardware to let the shop floor 'pick what worked' and used an arbitrary dollar value for costs.

We quickly learned that there was certain hardware that needed to be specified for application - suspension, steering, fastening into blind holes - and it was hard to get the message to the shop floor that sometimes you follow the drawing and sometimes you don't. The easiest resolution was to specify exactly what you wanted them to use.

For the use of parts having broader tolerance than specified, if they are truly functional, the source document requirement should be adjusted to reflect this.