Rivet length call-out on drawings 12

[JohnnyEnglish]

We don't call-out rivet lengths on drawings. We specify the rivet type as far as possible and then put XX for the length e.g. MS20615-4MXX. The quantity in the BOM is "as required", and we stock a selection of lengths that are reordered at a minimum quantity reorder point. Rivet lay-out is either determined by the company riveting standard procedure or by laser cut holes in flat patterns. The requirement is for the technician to select the correct grip length for the particular assembly.

Is this fairly standard practice on drawings? Any suggestions on other methods at other companies? What do the Boeings and Airbuses of the world do? How do you get around variations in grip length due to sheet metal thickness variations?

Thanks.

JE.

 

[YoungTurk]

Here (another OEM) we also do not call out rivet lengths on drawings. We do, however, have a process specification identified which contains a process to determine the correct length and inspection requirements for the finished product.

Even with screwed / bolted connections, where we do identify fastener lengths, our process specification allows change in length to account for variation (+1/-2 grip length).

 

[rb1957]

similar to above, we ("chop shop" = "pimp your plane") use A/R and lengths are determined from process spec and final stack up.

an OEM i worked at specifies the number of rivets, of each size and length (!!) ... they have a group of people to count rivets on the drawings (poor bastards). this is what happens when engineering (you buy rivets by the pound/kg) gives way to accounting/MBAs ... sigh

 

[KENAT]

Pretty sure my place used to specify the length, but we were on the fringes of hard core aerospace on the defense side.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[ncaero]

We (DoN) have fully caved to accountants/MBAs, and usually specify the exact # of rivets and grip lengths.

 

[tbuelna]

I have worked for Boeing and McDonnell-Douglas. Back in the good 'ol days when we actually used drawings, if I recall correctly the practice was to call out the basic part no. with material/finish code and diameter, but no grip length. The manufacturing process spec gave instructions to the mechanic on how to determine the correct grip length at installation. This allowed for variations in stack tolerances, shims, gap fillers, etc. I also recall that the specs often allowed for a longer rivet length to be cut down if the correct length rivet was not available. As others noted, the BOM qty was always A/R.

These days the big aircraft OEMs have mostly transitioned to model based definition (MBD), so there are no longer any drawings. Each rivet is just represented by a point in the digital model, embedded with information.

 

[WKTaylor]

my 2-cents worth...

Hand drawings and 2D-Cad drawings allow the use of drafting short-hand... allowing designers to avoid excessive detail for trivial elements such as solid rivet grip lengths.

However...

CATIA and other extreme 3D Cad packages mandate exact fastener definition... to the point of being anal. This makes for a great parts count... but can be tedious for everyone... and very expensive.

My philosophy.

When calling-out common soid rivets MS20470, MS20426, NAS1097 then extreme detail, such as length, is unecessary: shop should draw from common bin-stocks.

However, when calling out specialized rivets such as non-standard alloys [D, DD, K , KE, T], or precision spec parts [NAS9313, NAS9314, etc], or special oversize ODs [NAS1241, NAS1242 etc], or blinds, [etc], then length C/O is a detail not to be blown-off as trivial.



Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[YoungTurk]

CATIA and other extreme 3D Cad packages mandate exact fastener definition

Not true. You can still structure they data as you see fit. In our case we include the rivet head in the 3D model, which is sufficient to identify type, size, location, and orientation. The length, as well as required hole definition, is still per process spec.

Fastener quantities are included by type in the BOM in this case (automatically), though I'm not sure whether it is very helpful without the length.

Good point about non-standard fasteners though.

 

[drawoh]

wktaylor,

In SolidWorks, I create a hole pattern. I insert fasteners into the first hole. I create a pattern by clicking on the fasteners and the hole pattern. This makes it easy to manage high part counts, as long as I understand what fasteners go into the hole, and as long as everybody builds to the drawings. If someone is inserting shim plates, all bets are off.

Other 3D CAD packages should be similar.

My experience with rivets, admittedly on aircraft mounted "appliances" has been that I quote the diameter, head, and material, and I let the fabricator sort out the length. Nothing I have done with rivets comes anywhere close to the complexity of an aeroplane wing.

3D CAD, if it is well managed, can make it way easier to manage complex BOMs.



--
JHG

 

[JohnnyEnglish]

Thanks to all for your replies and discussion. I tend to agree with no length for standard solid rivets, but detailed call out for non-standard special hardware such as Cherries or Hi-loks.

So, assuming the rivet length is not specified on the drawings and left up to the technician at time of assembly (which makes a great deal of sense from many points of view including certification, design and production), how does your supply chain handle the ambiguity with respect to ordering and traceability?

 

[berkshire]

The old standard used to be to supply a segmented tray with a spring loaded set of lids to the Mechanic/Technician containing the type and style of fasteners required for that particular assembly. At the end of the task or shift the Mechanic returned the tray to stores, for refill with approved fasteners.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[Sparweb]

On a drawing, I only call out the length of the rivet if OUR shop will be making the parts. I know the guy doing the work (sometimes it's me) and I know this step will save him 32 seconds of checking, and he's the kind of person that likes that stuff. Secondly, I tend to know a lot more about what materials are available in our own shop, so the chain-reaction of various material substitutions is less of a mystery. A shim here or an up-sized sheet thickness there can throw all rivet length calcs out of whack. I know it's going on when it happens in our shop, but in an outisde client's shop I won't find out until after the build is complete.

Otherwise I don't bother, because our outside clients don't seem to care. Most have highly qualified people who have no trouble with this.

For the time being, we are not scrutinizing every last fastener's traceability, just noting the origin and batch, and checking things over at purchase time. We tend to order multiple lengths at the same time of any given fastener, so it's pretty easy to figure out what came from where, when that batch was used, and on what.

STF

 

[mfgenggear]

what happened to tolo's "mylar tooling layout" with the rivet hole pattern.
use to be what the mechanics match drilled with #40 pilot holes.

Mfgenggear

 

[YoungTurk]

What part of this conversation leads you to inquire about layout methods? Nothing about choice of CAD system mandates how production choses to tranfer layout to the aircraft. That said, layouts tend to be easier to create from 2D data since there is not necessarily a "flat pattern" available to convert to a layout with 3D data. On the other hand, 3D cad systems lend themselves to tasks like machining or projector layouts nicely.

 

[mfgenggear]

Youngturk

that how the engineering was transfered from the drawing to actual work done.
the center distance, edge distance, the qty , was transfered to to the hardware.
it was all held on a holding fixture then the guy would fasten the mylar to the parts.
hows it being done to day.

how does the 3d help here. It may be great for the designer how about the guy on the floor.
The data needs to be flowed down.

Mfgenggear

 

[KENAT]

What does this have to do with rivet length or am I missing something?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Sparweb]

It's a bit of a tangent, Kenat, but what they are getting at is the subject I was concentrating on in my posting: the rivet length is most valuable as a work instruction, for the guy on the floor, doing the work. Most engineers/designers/drafters will ignore the rivet length unless they are aware of a characteristic of the joint that makes the rivet's length matter (unusually long shank, subject to tension/prying, etc.).
Likewise, rivet patterns, spacing, size, and the process of transcribing all of that information from drawing to part is often left up to the shop to do 'somehow'. Actually there are standard practices that come into play, specs in company structures manuals, and lore and legend too.

So the subject seems to be drifting to how information is communicated from CAD station to shop floor, more generally... Not quite off topic yet, but not far. Thanks for keeping an eye on it.


STF

 

[mfgenggear]

Kenat

it's all related, The point I am making is 3d engineering is great technology,

But it's was up to the the guy on the floor to make sure it was assembled correctly.
It had to meet the OEM specification.

I am asking a question how does 3D help the guy on the floor.
may be I should posted on on an other post.
I leave it alone.

Mfgenggear

 

[tbuelna]

......how does your supply chain handle the ambiguity with respect to ordering and traceability?

JohnnyEnglish,

With bulk materials like rivets, most larger aerospace companies now use a "pull" type inventory system. I believe the process originated with Toyota in the automotive industry and they refer to it as "kanban". To order the rivet, the OEM's have Qualified Supplier Lists for every part number they purchase, and each part number has a source/vendor item control document or procurement standard. Rivets don't usually require traceability, but if they did the traceability would be by lot number.

Here's a brief description of the "kanban" process:

http://leanblitz.net/2012/05/toyota-way-principle-3-pull-systems/

If you want to learn more about configuration management and traceability in the aerospace/defense business, take a look at MIL-HDBK-61. It is 220 pages and covers the topic in excruciating detail:

http://www.everyspec.com/MIL-HDBK/MIL-HDBK-0001-0099/download.php?spec=MIL-HDBK-61A.015177.PDF

Hope that helps.
Terry

 

[2297Rocky]

I have to agree with most of the replies I have read on this thread.

I have worked on both military and corporate aircraft for years and everyone is different on rivet length call outs originating back in the day of hand drawings, most were rivet type and material call-outs, leaving the length blank or calling out an "XX".

Rivet counts were mostly A/R.

When BOMs moved from the first page of an assembly drawing to a "Parts List" it would reference a drawing zone, but would still have and A/R for most common fasteners. Only in the case of special fasteners, was there ever a count called out and then usually the length was still omitted.

With respect to larger companies there usually is a cryptic code called out for a particular fastener, which may be an NAS, MS, or AN fastener and sometimes may have a company part number attached to it as well.

Corporate and commercial are good at assigning there own part number which may have and equivalent called out to a Cherry or Huck part number or may be a modified and proprietary part.

IPCs are wonderful, but usually will not go that far as to call out a fastener, especially a length, that's usually in the production drawings which may not even be available, however military tech manuals are a little more in depth, usually because they are 4 or 5 inches thick and normally more than three volumes to each tech order, but when it comes to length, there again it is usually the guy on the floor that will make the final decision for the application.

If it were me, I wouldn't call out the length unless I was 100% sure of the stack-up especially in bonded assemblies where the adhesive bond-line can vary several thousands and when you throw in a filler, a doubler, tripler, and a couple of skins and a variance due to several layers of adhesive.

If your company training and processes are in check, he or she will make the correct selection and procurement will love you all for it.

Best Reason: call out the length of a fastener which costs a couple dollars and it usually is no big deal, call out a fastener which runs around the c-note per each and there are several hundred of these high dollar fasteners in you assembly and you could have a VERY BAD DAY real quick!

All of this and we still go to work each day? Go figure!