What does your shop do? Input please. 2

[NFV1]

I am a design engineer working in a manufacturing machine shop that employs around 40 machine operators (most of whom don't have a great deal of training and have no machine programming experience). As it's not my game I don't get involved in manufacturing engineering, yet I am constantly told "we can't do that" or "we can't hold that tolerance". Then on top of that almost a forth of all parts are out of spec one way or another. I understand that often there is a riff between design and manufacturing engineering. As I don't have a great deal of experience in other manufacturing plants I am looking to see what the "standard" practices are regarding machine shop manufacturing practices.

Do your companies...
1.Supply a job tray with ALL necessary equipment for particular job?
2.Give operator a program and rough route in order to complete part?
3.How accurate are your routes? 100%, 75%, 50%, 25%, never; and how important are these to your company?
4.Are operators required to manipulate programs for slight changes in final dimensions rather than making a bespoke program?
5.Can you add multiple routes for the same parts (say making part of load center 1 vs. load center 2)?
6.How much special set up equipment (ie jigs, arbors, and the like)do you use?
7.How much time does a manufacturing engineer spend on the floor.

a)Run ever part the first time with operator?​

b)Never run first parts with operator but available if issues arise?​

c)Never on floor (this is operator and floor supervisions responsibility?​

Any input would be very helpful. Thanks in advance.

 

[JNieman]

Standard practice is to have a manager who is smart enough to call them on lies when appropriate, or motivate them to achieve better when they should. Your machinists and operators can be your best asset in creating intelligent, efficient, and cheap-to-make designs, but only if you have a good relationship with them.

1- That should already be happening. If they really are just operators, you should be supplying most of their requirements at least. If they are machinists, it's common practice in many industries that the machinist is required to supply non-specialty tools (not consumables) but that is sometimes frowned upon in other environments.
2- Programming is dependent upon company practice, type of work, and employee resources. We actually split it. Most machinery is programmed in the office and a 'setup sheet' is given to the machinist to set up and run the parts. Other departments program "at the machine" and are given nothing more than the part requirements (drawing, usually).
3- I don't understand the question.
4- Final dimensions are set - that's the requirement of the part. Period. If design changes need to be made, they can be suggested by -anyone- in the company, but the designer (or the designer's boss..) is the final say on whether or not it's a positive change.
5- I don't know if I understand this question either, but it may be because I don't know what you mean by 'route'.
6- A lot.
7- a- Some operators are quite adequate at setting up their own machines. It's a case-specific decision.

b- Same as above.​

c- Here, we don't have someone who is "just a manufacturing engineer" but instead also wears multiple hats. CAM programming, Quality management, tooling/fixturing design, and so time on the floor is hard for me to pin down. Some things require more first-hand observation than others, and sometimes it's just convenient to do it that way even if it's not /mandatory/ so we do. For high quantity runs, the answer would be "never, unless there's a problem that requires input".​

There are typically two situations that yield your problem -

1- the designer has no idea what the "appropriate" capabilities of the machinery/process is, and thinks they can get a mirror finish off a bandsaw.
2- the "tail is wagging the dog" and management doesn't know when they're being misled by operators/machinists that are being lazy or are scared to commit to better results.​

The reality is often not as harshly as I put it, but I use strong words for the sake of being clear and direct.

 

[NFV1]

JNieman thanks for the input.

Clarification:
"Method of manufacture" is referred to as a "route". This is the process map of the part and is leveraged by accounting for cost and manufacturing for process.

Regarding the "typical two situations":
1- As a designer I often send R&D to vendors and get fantastic results, therefore I feel our designs are achievable. I should also note that I'm typically impressed with the questions and diligence paid by the vendors.

2- I don't feel it's the operators or machinist but rather the manufacturing engineers/programmers that are setting up the operators to fail. Not giving them what they need to be successful. Often telling them to not follow print or to not follow the the specified routing. For instance telling operators and QC by word of mouth that a ID spec'd to 7.210-7.220" should be .010" over top spec. I'm assuming this happened because they had an issue fitting the gauge. Then the designer hears about this from QC. It's so bad that we have a term for it... "Tribal Knowledge". I completely appreciate your response to #4.

Side notes on the shop:
-Before me the one design engineer worked off location and didn't see the day to day. He simply designed and sent drawings and layouts. Now we have two in house design engineers.
-Most of the "manufacturing engineers" are older and seem to be against any change (some won't even use 3D modeling). Also most are promoted machinist.

 

[JNieman]

We have our share of 'tribal knowledge' as well. Don't feel like your situation is very exceptional. Finding out why they feel they can/should deviate from the print is important, though. If there's "no good reason" then it should be corrected. If the parts don't work /unless/ they deviate from the print, then the design must be revised. Or maybe the mating parts are never made to spec. Making a bad part to offset another bad part is not unheard of but it's not ideal and it's never good to see parts made out-of-spec intentionally. One must be made to match the other, whichever way it goes. Otherwise, what happens when you give the job to a new "tribe member" who didn't hear about that "tribal knowledge" you speak of, yet? Now he's making good parts (to print) that aren't good (per the tribe) and no one was really at fault, and thus cannot be corrected, because the system was broken.

I speak from experience, unfortunately

I understand what you mean by routing too. It's important to follow the routing for reasons the shop often doesn't understand. We have waterjet and CNC mills, for example, and I can think of one example where someone on the shop decided the parts were "too tight of a tolerance" to be made on the waterjet, and brought them up to go through the CNC mill department. Then we lost a lot of money on that job. That job was quoted at waterjet rates. Not programming+VMC rates. There was nothing that couldn't have been done on the waterjet. Someone just made a bad decision. I personally went and programmed, set up, and ran a sample part from falloff/scrap material just to prove the situation, and I'm not even the guy who sets up and runs the waterjet. I'm just a guy who decided to learn it as a backup. I'm the "back up" guy in case of emergency. The primary people should be better than me. They are. It was just a bad decision. The tail wagged the dog, in that case.

So changing processes is often problematic. There are myriad reasons one may not know the impacts of changing processes either. Sometimes it's money. Sometimes it's because it effects downstream processes.

It's hard to shed the "small shop" mentality as you grow into a "mid sized shop" because "that's how we always did it" and "word of mouth" work well in a small shop when you can yell 20' to the next guy. That doesn't work so well once a company reaches a certain size (both in # of employees and physically) and experiences turnover (retirement, firings, people moving on, dying, etc)

 

[Skyeman]

Jamie (Aerospace & Automotive)
It might help if you indicated what geographic area you’re in and what customer base you serve?
(Seattle tool shop may not really be comparable to one outside of Jo’burg serving a tractor plant)

First, as a general remark and without knowledge of your customer base, if 25% of your parts are out-of-spec then you need to get your CEO involved before the business goes pop!

Second, the cooperation between Design- and Manufacturing Engineering should be so good that there are almost never hard-to-resolve disagreements. The design people (again without knowing your business) should be VERY close to your customers and should appreciate the consequences of dimensional variations on subsequent performance of your part or assembly in the ultimate product.
Your sales people should know the business, top to bottom; if there is any doubt at all they should be replaced and/or re-trained and they should involve design or R&D in their pitching-for-business meetings whenever that’s required.

All of the above is pretty standard and your CEO should have it all in place already.

Your Qs:
1. It depends. In many cases you will have to supply all required items – this is especially true if you are all-CNC. If you also include some manual machines then some of the decisions can be left to the operators – examples might be using a roughing turning tool with / without chipbreaker.
2. JNieman answered this. Regarding the route, a lot of jobs define their own route – If you’re doing ultra-precision parts, for example, you might have to machine a bore with 3 passes ; roughing, semi-finishing and finishing and that could involve more than one machine, a logical sequence. The manufacturing people should know the strengths and weaknesses of their machine park – one machine might have poor spindle bearings, for example, making it unsuitable for anything more accurate than roughing, or semi-finishing. Again, any manual machining routing can (within limits) generally be left up to the operator.
3. See (2) above.
4. ANY change to a finished, functioning and „signed-off“ program should be forbidden without the agreement (normally covered by a Change Note which is recorded and then forms part of your internal- and customer quality tracking system). Generally the change route is through the manufacturing engineer to the designer and then ultimately to the customer.
It seems that this route is broken in your company; this is very serious and demands attention.
5. See (2) above which touches on this. Your manufacturing people MUST be on top of their equipment capabilities. The general answer should be Yes but the alternative routes must be spelled out for the shop floor.
6. A lot!
7. Some manufacturing engineers are actually jig-and-tool designers quite a lot of the time and inspectors/QC at others. The general answer is: as much time as is needed to ensure the shop is ticking alone nicely and producing quality stuff.
A walk around the shop by an experienced and knowledgeable engineer can often pick up issues before they become serious. If your people are on a bonus system then the “I’m watching you” negative aspect of this has to be taken into account – the ManEng should be seen as a helper not as a hindrance to making money. If the 25% out-of-spec is happening frequently then it seems a lot of internal re-training, talking about roles, honing skills, etc., needs to be happening.
If, as you indicate, most operators don’t have much training, then I would say the first-run part(s) should be done „together“ – with highlighting of dangers, reasons for, etc., being pointed out and discussed openly.

As JNieman indicated, the Tribal Knowledge thing – if left unaddressed – may ultimately kill your business. There are positives but the negatives greatly outweigh these in a modern shop using off-machine programming, CAD-CAM, 3-D modelling, etc.
(Get your CEO to investigate 3-D printing!)