Design for Manufacturability & Assembly at low production volumes 2

[KENAT]

We had 3 days of DFM/A training last week and frankly I was a bit underwhelmed.

1st off in terms of actual things to do or not do I didn't learn much I didn't already know. There were a few things but nothing outstanding.

2nd it seemed tailored to much higher volumes than we deal with. Don't get me wrong, I'm not saying DFM/A should be ignored for lower volumes but it seems that a lot of the trite truisms that get spouted are less applicable. Also the time spent on the DFM/A analysis is a lot more significant amortized across a few dozen parts compared to hundreds of thousands.

However, manufacturing engineering management claim they are serious about it so I figured I'll pretend to drink the kool aid and go along with it.

So, in that vain vein, can anyone point me to any resources that really address DFMA/Concurrent Engineering/DFX or whatever you choose to call it for low volume production? I did searches on here & Google for many of those terms bud didn't turn up much.

To give you some idea of volumes, our highest would be somewhere under 300 units per year for one of our electrical control boxes that get used on multiple product lines. Our more popular actual microscopes sell perhaps 75-150 per year. Our less popular ones perhaps less than 40. Our really big automated systems and various accessories may sell 5-15 a year and we also do 'specials' which can be one offs. Product life is typically under 10 years so even our highest volume control boxes would be lucky to hit total production much more than a couple of thousand.

Thanks for any help.

(I know this should probably go in "Industrial/Mfg engineering other topics" but that forum doesn't get much traffic so I thought I'd stand a better chance here.)

Posting guidelines faq731-376

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

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

 

[MikeHalloran]

DFMA is _real_ big on snap fits. ... which you can do in low volume plastic molded parts, and to some extent in sheet metal.

Another thing to look at is self-drilling and/or self-tapping screws, and especially, severely reducing the actual number of screws. ... e.g. by using drive rivets or cable ties instead, and/or by combining parts.

Like you, I was sort of underwhelmed by B-D's presentation, mostly because I already knew about all the stuff they recommended. My then employer paid for training for all the engineering personnel, and bought one station of the software. ... which almost nobody used thereafter, because the time to use it wasn't written into anyone's project schedule or budget.

Our primary justification for not using DFMA techniques was serviceability. We were well known in the marketplace because our service people were on site nearly every week. ... and we eventually lost much of our market share because our service people were on site nearly every week. We couldn't find a way to build our stuff well enough that it never needed disassembly, so we tended to use fasteners that would survive repeated reconstruction.

Our competitors made their major modules cheap enough to throw away when they broke, and they didn't break very often, whereas ours were always suffering from stupid design errors.

... which existed to a large extent because it was an 'electronics' company, and our electronics wizards exerted way too much influence on mechanical decisions, which they regarded as relatively trivial because everyone could see the mechanical stuff, so it couldn't be that hard, etc. We were plagued by undersized bearings, overheating, excessive deflection, bad material selection, stupid/bizarre fluid systems, etc.

DFMA might have made a difference, IF the company were willing to put some actual time into mechanical design and engineering up front. ... but they never really had. Their 'best' mechanical designer was real fast, because he hardly ever erased anything. If you pointed out that a bracket needed to have another function, he'd add another bracket instead of rethinking the errant one.

... and so on. They got what they paid for.


Mike Halloran
Pembroke Pines, FL, USA

 

[jmw]

Management is so obsessed with mass production, standardised, straight from the box, fit and forget plug and play, even when there is no good reason and even in niche market industries where the very fact of being niche market means cost is less of an issue.

If only you could get them to spend money on what you really want, need and would make a difference.

I spent a lot of time working out how to make a certain product work in the field (my job description even included all those clichéd terms, and discovered that some (most) applications are so critically dependent on process conditions that you had to custom design each system.
If you tried to peddle a factory standard solution you ended up with an expensive pile of junk and a big after sales service cost.
Fortunately the system integrators accept this and just buy the standard sensor and design the system around it themselves.
I am talking about a primary sensor that sold by itself can sell in thousands at reasonable prices. But that is thousands in a market with a potential for millions and where to get that market you do have to drive costs way down.
I am sometimes surprised by the price spread.
The price of a single measurement can vary from, in the millions market $50-$60 (no one seems to have got there yet with a suitable sensor.... I nearly got one manufacturer there but since it included the US automobile industry he didn't want to play having been ripped off there before) to an industry where they happily will pay over $1,000,000.00 installed cost for a single online measurement with payback very conservatively estimated at less than a year (and with an ROI demonstrated by some users as a couple of months).
And this was the product and application for which my management wanted me to deliver a standard off the shelf solution.

Management logic is often as incomprehensible as that of that other species we coexist with, one of which now wants me to take her shopping.

JMW

http://www.ViscoAnalyser.com

 

[mcgyvr]

DMFA is something that every engineer should have been doing long before this "DMFA" buzzword crap came around. I just call it good engineering/design. During the whole design process I'm always thinking about "how will this effect assembly, will additional tooling/training be required, could I make this in 1 piece versus 2, do I really need that screw, can I get rid of that rivet, etc.."

Like you eluded to already its pointless to do "DMFA" just to remove 1 screw from an assembly but it took you 200 hours to figure that out and now you added "x" amount of dollars to each piece because of time spent. So yes its tougher to design something properly for low volume work.. I do it all day though..We are a high mix/low volume shop. But when you know what you are doing you tend to build it the best you can the first time around.

I just can't stand the term DMFA and how its such a huge thing in the industry. It just means I didn't do my job right the first time. I read an article on it the other day and they showed this bracket assembly made of 4 pieces/rivets,etc.. and how their "DFMA" expert took it down to 1 piece saving X amount of dollars.. I saw it as.."The engineer was just an idiot in the first place and any monkey that knew 2 cents about sheet metal fabrication would have NEVER made that bracket in 4 pieces"

 

[SnTMan]

Without fads and buzz words, how would consultants make a living?

Regards,

Mike

 

[drawoh]

KENAT,

I have not done the training, but I have the Boorhroyd Dewhurst book here. It was a nice read. I am a technologist with a three year diploma, and I took a lot of courses on manufacturing processes. I am under the impression that this is not taught at the university level. Mechanical designers need to understand this stuff, and the B-D book makes up for this to some extent.

What you are doing is very similar to what I am doing. We make electronics stuff in quantities of one or two, to up to forty units a year. Our production is interested in DFMA. Most of engineering is not showing interest. We did design a system where it took a pair of scientists a couple of days to put a cover on, so it would help us to think about it.

I can confirm MikeHalloran's experience with electronics people. Everybody thinks they can do mechanical design.

Several of our worst assembly nightmares were delivered to production as stacks of fabrication drawings, without assemblies and parts lists. I am convinced that a big part of DFMA is documentation. Creating the document forces you to think about it. If the mechanical designer brings a complete assembly drawing to the design review, you can review the assembly procedure. You can make sure it is feasible, if not efficient. You can identify the need for alignment and assembly tools, and integrate these with the design.

You could build an enclosed, heated patch of grass next to the front door where we could park our unicorns.

JHG

 

[mcgyvr]

SnTMan..EXACTLY

Many times I've wanted to just go into companies and tell them that I will look at every aspect of their company and provided solutions to save them money.. I do not want to be payed a regular salary at all..If I don't save you money you don't pay me no matter how much time I spend there.. But what I do want is simply 1% of the money I do save that company.. I'd be RICH.

 

[tygerdawg]

There are probably dozens of texts from a variety of sources that you could ponder upon and digest the info. If you're old enough and still employed doing design work, I would guess that you probably already know most of what is in those books. Common sense, really: unidirectional insertions, snap fits, intelligent fasteners, asymmetrical orientations, bosses, yadda yadda yadda.

My professor in grad school was a Boothroyd-Dewhurst advocate, and even so we all felt we learned the foundation of thought in about a week. Fast forward to present day, the thing I see missing in design education is the concept of "practical" or "sufficient" tolerancing. I learned this during my exposure to Shainin Six Sigma training. This concept bridges the gap between the designer setting tolerances on paper based on his knowledge & experience, the manufacturer trying to accommodate those tolerances based onequipment capability, and what is really necessary to make the assembly do the desired function. In the Six Sigma world of non-obvious parameter interactions, those three items don't always intersect or agree, and the consequences of not looking into them can cause a lot of unnecessary heartache.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[MintJulep]

For low volume stuff there is obviously less opportunity for actual useful advantages, but there are still some things that make sense.

Establish and use a small selection of preferred fasteners. Head types, diameters and lengths. This reduces the need for tool changes, the likelihood of using the wrong fasteners and time spent searching for the right fastener, stocking etc.

Use electrical connectors, terminals and such from a single product line to reduce the need for different crimpers, pin inserters, etc.

When possible, use a common design and size for access covers, panels, hinges, latches, etc.

Look for easy ways to reduce part count. How many stick-on or rivet-on labels do you attach to something? Could they be combined into a single label? Could labels be replaced by silk screen, laser etching, having the information molded into plastic components, etc.

 

[jmw]

Reducing part count doesn't always result in a better or cheaper product of course.
The classic example is a pair of scissors.
At one time just three parts, the two mating blades and the pivot (screw, rivet whatever).
Now scissors can have five parts maintain quality and be cheaper.
The blades, the pivot and two plastic handles moulded onto the blades.


JMW

http://www.ViscoAnalyser.com

 

[HDS]

A big part of DFMA at any volume level is doing the tolerance stack-ups to make sure things will allways fit together and if possible not require adjustment. The trick with low volume is actually getting the parts made to spec...

I your quest for literature look at aerospace DFMA texts.

 

[KENAT]

Well, thanks all.

Nothing really jumping out at me, the aerospace idea might have merit - any particular texts you'd suggest HDS?

I was trying to stay open minded and not be a 'negative Nelly' but pretty much every post above confirms my opinions etc.

I had someone tell me yesterday I probably could have taught the class - which is a bit of an exaggeration. However, I probably could have taught the actual 'design' portion - it was the mind numbingly thorough & long winded analysis methodology that was new.

Posting guidelines faq731-376

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

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

 

[jmw]

I do hope you gave some constructive feedback to the trainer or trainers company.
This is probably a standard presentation given every other day to the same old companies.
Then they start to look at smaller companies looking for the bottom end of their market.
That could mean they really haven't thought through their presentation and haven't maybe understood the differences between yours and their usual client runs. It could be that there was something to learn and some benefits to you and they failed to connect because they failed to properly understand the more niche end of the market.

JMW

http://www.ViscoAnalyser.com

 

[KENAT]

Well, not sure quite how constructive he may have considered it but I definitely gave feedback - I tried to actively participate asking questions as we went along etc. despite my gut reaction being to just sit there and endure it.

For instance, after 20 or so slides on design for injection molding plastics I did ask why he decided to present that when earlier in his own material he'd already suggested methods such as thermoforming may be better suited to our volumes and our own experience backs this up. The answer was essentially this was part of his standard presentation.

Also after a similar presentation on casting I did point out that we'd already identified and used Rubber Plaster Molding which actually mitigates a lot of the typical drafting and wall thickness variation concerns etc. which his slides focused on - as well as being suited to our volumes.

In fact except for adding half a dozen photo's of 'how not to do things' from our products into his notes and condensing the presentations length to meet our schedule I saw little evidence that he'd even tailored the presentation to low volume situations, let alone our even more specific needs. Of course, this may be due in part to what our Manufacturing Eng Director asked him to do/what info he was given.

Posting guidelines faq731-376

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

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

 

[BillPSU]

DFMA includes a great deal of items besides snap fits and sheet metal. As a designer do you have the list of standard materials which the shop currently purchases. Design using these materials will make it a lot simpler to get the one off parts. New materials may be needed in some applications but use standard materials if possible. Using currently available fasteners if possible.

Try making a part made from 16" OD x 4" wall x 18" length tubing. We had a saw which could cut 18" material and we had a 22" swing CNC lathe, no problem right? Wrong, the hoist above the saw had one ton capacity. Try picking up the bar which weighs 415 lbs/ft x 20 ft with a one ton hoist. We bought the part sawed to length. Design engineering had to create another part number for the cut to length piece.

Most DFMA issues are using common sense but inexperienced designers need coaching on using it.

 

[MintJulep]

However, manufacturing engineering management claim they are serious about it

So

Have they established a quantifiable goal for the exercise?

Analysis showing that the goal is realistic and achievable?

Metrics on how to measure progress towards that goal?

A schedule for implementation?

Processes to be followed?

Have they decided who will be to blame if (when) it doesn't work out how they imagined?

Or is it "Well, we trained you. Why haven't we saved all the money that we thought we would?"

 

[KENAT]

Well Mint, I pretty much opened up the training with the career limiting question about if management were serious about it this time and if they were willing to put the time & resource etc. into it... and was told yes and the fact 25 or so of us were taking 3 days out to do the training was a sign of this...

Then the training focused a lot on how it did mean doing more up front, and manufacturing engineering taking a lead in doing the DFMA analysis etc.

Then toward the end of the second day where we'd suggested the emperor was naked for the 10th or so time as we'd never be allowed to spend the time up front... the Manufacturing director basically said well now you design folk have been trained we won't need manufacturing engineering to do the analysis or even to do the analysis etc. and so basically ignoring a major theme of the training - and the fact that many of the engineering folk knew most of the actual things to do or not do long before the training just weren't usually told/allowed/incentiveized to make it a priority compared to program schedule & product performance.

Anyway, I didn't' mean this thread to turn into a rant about management fads or consultant schemes etc. I really was trying to be proactive about finding information targeted to our volumes - including simplified ways of doing the analysis etc. when you have less volume to amortize the inherent costs across.

Posting guidelines faq731-376

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

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

 

[TheTick]

DFM = design for manufacturer, i.e. design to be made with a given process.

Low volume means more agile processes (e.g. CNC), fewer mass-production processes like molding and casting.

THEREFORE...

Design to the processes that lend themselves to agility and low volume.

BOOM.

(Note, for added credibility, the proper employment of the terms "e.g." and "i.e.")

 

[KENAT]

(By the way drawoh, love the unicorn comment, just one question - how far can the heated grass area be from the the rainbow's end and are there any clearance issues with the flying carpet take off zone?)

Posting guidelines faq731-376

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

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

 

[drawoh]

Well Mint, I pretty much opened up the training with the career limiting question about if management were serious about it this time and if they were willing to put the time & resource etc. into it... and was told yes and the fact 25 or so of us were taking 3 days out to do the training was a sign of this...

As far as I am concerned, production requires more design time. Even over a total production run of fifty or a hundred pieces, an day saved in design time can translate to many, many days of production time, and pissed off customers. For an anticipated run of a hundred pieces, I think you should take two or three designs to the point of assembly drawings and some crude, preliminary fabrication drawings. You can evaluate part costs and assembly procedures. You can evaluate the product's applicability to your market. DFMA is part of this.

You need management to commit to the time. If your in-house control freak insists on having a plate made, then another plate and then a bracket, etc., then there will be no benefit from DFMA.

In limited production, you do more machining and sheet metal, and no injection moulding. Boothroyd, Dewhurst and Knight cover these processes in their book. As I noted above, the book probably provides useful knowledge to someone who has not been taught anything about manufacturing processes.

You appear to respect the person I ripped off the unicorn comment from. Pots of gold attract leprecauns. This might not be safe for the unicorns.

JHG