References for designing and quoting Job Shop work.

I work in a job shop that sees just about anything imaginable come through the door to be quoted. Much of it involves single parts that the customer wants modified,fixed or rebuilt. I have only recently moved into the role of quoting these walk-in jobs. I don't have too much problems quoting machine times close enough. But I do have problems deciding what methods and design are best. For example; I have plenty of manual machining experience making accurate parts, but not such a broad knowledge base when it comes to choosing the best type of bearing to use, whether or not to use a dowel pin vs. a roll pin, if a press fit is sufficient or does it need to be welded, when to use splines vs. key and keyway, when should a bolt pattern have a few large bolts or many small bolts?, etc.
I realize most of this knowledge comes from experience of seeing a lot of different part configurations; both successful and failing. I don't have illusions of learning this the cheap, quick and easy way, however, I have to think that there's got to be some reference books, pamphlets, videos, etc., that cover some basic elements (some folksy rules-of-thumb) of part design that help the quoting/design process without involving mind-numbing engineering formulas and extensive trial and error. Most of our customers aren't interested in paying for hours of engineering research in addition to machine time.
Does anybody know any good manuals, reference books, or websites?

I appreciate any leads.

My brother and I have had a small jobbing shop for over 40 years and I have yet to come up with a foolproof method for
quoting one-offs and design heavy jobs. A lot of it is just leaning on your experience and, yes, even going with your gut
instinct.

If your customer can't tell you exactly what they want--we run into this all the time--then you have to design it in the best
way that you know. Often you have to go to consult with other people in the shop and, of course, there are lots of outside
resources as well. Call a bearing supplier and let them help you with the selection of parts; call a powertain specialist if
you're stuck trying to figure a motor/gearbox combination.

A lot depends on your relationship with the customer. We have longstanding customers where the relationship is such that
I'm often comfortable with a simple hand-drawn sketch--we know each other well enough that we can end up with a good
finished product without a lot of fuss and bother. Other customers are just too darn fussy to take that chance. I want
something on paper from them before I'll tackle a project--experience has taught me that they really don't know what they
want and that there's a real good chance the design parameters will change halfway through the job.

The only time I really "quote" stuff like this is if we're short of work and even then I'm pretty cautious. I do lots of "estimates"
to give people an idea of what they're going to be spending but I want to have a pretty detailed design if I'm going to lock
into a price....

If you're not going to get an engineer's training, you basically have to see (work on) a lot of stuff that HAS been engineered, in order to become familiar with how stuff is made. Then you adapt and/or copy those principles. So the stuff you work on becomes the same sort of stuff as you'll make.

If you don't have familiarity with a certain field, then you have to rely on the customer to either teach you, or he can get his own engineer to design it: you (as a machinist) can't do, or be responsible for everything, except making sure that the parts you make go together in proper assemblies.

The applications of various types of bearings are fairly straightforward from the design of their load carrying ability and the direction in which that design accepts thrust or loading. I also stress as much as possible that having to put things together, and take them apart yourself is great training in how to make choices in bearing fits. You learn how difficult it is take apart something that wasn't too bad to assemble because of too many tight fits, and then you start to LEARN how things work in the real world.

Conversely, in the repair biz, you learn why sloppy tolerances are bad in the wrong places: loose fitting sheaves, hubs, bearings all wear out their journals/keys/splines in short order when things are sloppy.

Splines are used to maximize the torque transmitting capability of a shaft that, for its size, is working hard. A hard working key puts a lot of stress into one side of a shaft, the groove itself being a stress riser. All the torque has to go against that one poor little groove, and it can initiate a crack more readily than the same torque being applied to a half a dozen grooves (splines).

Splines also are preferred when powered components need to slide under load, and that becomes obvious from the application. Splines are always preferred over keys (I would guess), but the cost of production limits their application to situations that justify it.

Press fit versus welding: as a general principle, you don't rely on a press fit when the load is in the same direction as the press fit is made. You just never know if the loaded member could receive greater loading than your press fit and then it moves and goes all to hell. The exception would be if the press fit is mechanically contained (by shoulders, nuts, etc) that ensure it cannot move further.

IMO, welded fits are necessary in some areas, but you need to be aware of the effects of welding on the metals at hand, and the inevitable effects of distortion. So you plan around these shortcomings by welding what needs welding, before you do the machining, as much as possible, so that you can remove the distortion that way. Again, you need to be mindful of the stress-riser effect of welding in the middle of a shaft receiving torque, or one that receives heavy transverse loading (conveyors). And you need to be mindful of what welding procedures to use on various grades of steel that are involved. Having to weld stuff together often dictates the choice of what you start out with (has to be weldable) as material. Then low alloy materials means lower strength, and so you have to beef up the part a bit because of that.

Welding or not depends somewhat on how easy the components are to handle in the field: will someone have to service this thing? Is it too heavy and cumbersome to be a welded unit? Is it possible to get the thing out after it is welded together in the final assembly? Overcoming some of these problems are why bolt flanges get used.

Neighbors weld shop get's this kind of stuff as well.

Usually the customer will get something workable
(a repaired, beefed up excavator part etc.)

But once in awhile you get stuck. As did my neighbor.

For example, a large company wanted a aluminum
hood for some process, and it was very large,
winched up/down all day long, and people would be
working under it.

So, working with the customer, it was agreed this shop
would come up with something workable, and then
run it thru a professional engineer, to get things
tweaked for loads, safety factors, etc.

And the P.E. issued a drawing documenting it, and then
more importantly, stamped it.

Edit: Your company's insurance policy should be referred to as to what you
can & cannot do in these situations.

HuFlungDung said:

If you're not going to get an engineer's training, you basically have to see (work on) a lot of stuff that HAS been engineered, in order to become familiar with how stuff is made. Then you adapt and/or copy those principles.

Click to expand...

That really is the long and short of it. If you don't have the education, you need someone with the education to learn from. The more of a partial education you have, the more you'll be able to self-teach from known-good examples. However, there's a big gray area where you "don't know what you don't know" and can't really understand whether or not you'll be able to find all the key characteristics in a basic machine in order to duplicate it.

Engineers design machines, but they also design them to use a little material/cost as possible while still being strong-enough / functional-enough / economical-enough. You aren't an engineer and so you cannot do that very well. You can overkill the things you're aware of, but don't fully understand. That's safe-enough in most cases. You can't overkill the things you never knew existed. That's when you get into trouble.

You need experience or examples to learn from. You can start reading basic machine design books, maybe. Shigley's Standard Handbook of Machine Design / Shigley's Mechanical Engineering Design. All good practical stuff. It can teach you how to solve questions. It can't teach you how to ask the right questions, though. That comes from experience.

Graphs and standard general applications are your friend for this sort of thing. Nomograms too if such can still be found.

With a bit of creative thinking it should be possible to find standard good practice rules for things like bolt up joints including the difference between sealing and load carrying ones.

Graphs for things like spline v key load carrying ability, recommended shaft sizes, roll pin v spirol v dowel and so on. This sort of thing you need a decent factor of safety so graphs are avery good way of seeing how close to teh edge your design is. Nomograms are considered old hat but its so much easier to see the effect of changes when you merely have to swivel a ruler instead of sorting tabulated calculations.

Textbooks tend to be simultaneously too deep and too shallow. Heavy on the maths and fundamentals but light on practical implications.

Pity the old Caxtons, Newnes et al series from the 50's and 60's didn't cover design and drawing office practice. I suspect such would have given a decent overview in sufficient detail to either be used directly for simpler stuff or as a starting point to drill down when needed. Kempe's Engineers Year-Book tried but I always found them too general. Great for letting you know what was out there and in what sizes along with decent how it works introductions with important points to consider but little in the way of applications and strength data. All gone now of course.

Internet seems to have the data but its very can't see the wood for the trees. http://www.roymech.co.uk is a useful resource produced by the late Roy Beardmore. Less daunting then the full blown handbooks. Much easier to get a handle on something you don't really know about.

I found the best approach is to do some serious homework with decent notes when you hit something likely to recur. Initially more work but worth it next time round. Doing just enough to resolve the specific problem meant I could never recall what I'd done and had to start over every time.

Clive

Will be watching this thread with interest as I struggle with some of this myself.

Regards.

Mike

Two words: Industry Standard.

I'm an experienced machine designer (BSME) and I've designed machines for just about every different industry - the only constant is, you work to the Industry Standard when the customer doesn't provide the information.

The Packaging world was as cheap and easy as you could make it, but I would see machines come back to us for retrofits from 20 years of service in a paper mill, covered in dust, dirt, and grime, and still be in perfect working order. Never saw dowels used. NFPA cylinders, three position cylinders, and cylinders with servo valves used instead of servos, because accuracy just wasn't that important. Constant drive motors with friction clutches over VFDs. Oversize everything, lots of bent and welded steel, heavy, heavy tube frames. I once saw a casepacker get hit by a towmotor hard enough to break the lag bolts holding it to the floor...and it never even stopped running. The working tolerance for the machine builders, who laid out and drilled a lot of holes with tape measures and hand drills, was +/- 1/16".

I used to design injection molds for a world-class molding company. 16 cavity family mold was $1.3 Million. A loose tolerance was +/-.005", and even those were held to +/-.002" or better, as a matter of principal. The real work was in tenths on 24" blocks, and getting inserts to fit in a way I still can't believe.

What was expected in the mold shop (c'bores and tapped holes for SHCS) would get you laughed at in the packaging world (line it up, drill it through by hand, and stick a nut and bolt through it.)

I have spent the most time, and still currently, design Automation equipment. First for an Assembly Automation company that served the Pharmaceutical, Green Energy, Automotive, and film industries, now as a platform designer for a machine producer. The automotive world has one set of expectation, and pharmaceutical is complete different from that. Green energy doesn't know what they want, and film always thinks they need tighter tolerances than they really do. The vast majority of these machines are designed in roughly the same way, but dowels are consistent across all for any touch tooling, but preferred mechanisms change quite a bit. For some customers, lasering a slot into sheet metal for a couple of bolts as an adjustment mechanism is fine, for others, it needs to be a milled slot for shoulder bolts, and for others it needs to be mounted on bearings, or a sliding fit between two milled pieces. Parts that have raw, torch-cut edges can share space on a machine with parts that were wire EDMed. Sometimes a tabletop needs to be Blanchard ground steel, flat to within .005" per foot, other times stock MIC-6 jig plate works, and other times it can be a wooden work bench with crap screwed onto it - all in the same system.

When there is no customer mandate, and no prevailing functional requirement for the design, Industry Standard trumps all.

As an aside: for the vast majority of what I do, weight is not a factor (I've never sent anything into space) and, in the grand scheme of things, steel is cheap. Much cheaper than having to redesign something, or God forbid, redesign and remake something on the fly because the original failed at a critical time. So while there is a thing as too much overkill, if there's ever any question on if that 1" steel part will be strong enough, just make it 1.5". You'll thank me.

And now that I'm really rambling on: keep in mind that your time has value. How many hours do you want to 'spend' at your shop rate trying to figure out how to make a one-off part $20 cheaper? For that matter, how many days do you want to spend designing something, when you could otherwise be making money doing what you're really good at?

*Edit* I hate myself for the self-promotion, but I'm trying to be better about 'putting myself out there': if you ever want professional design help for anything like this, feel free to reach out to me. It's what I do. Now please don't hate me

Johnny SolidWorks said:

As an aside: for the vast majority of what I do, weight is not a factor (I've never sent anything into space) and, in the grand scheme of things, steel is cheap. Much cheaper than having to redesign something, or God forbid, redesign and remake something on the fly because the original failed at a critical time. So while there is a thing as too much overkill, if there's ever any question on if that 1" steel part will be strong enough, just make it 1.5". You'll thank me.

And now that I'm really rambling on: keep in mind that your time has value. How many hours do you want to 'spend' at your shop rate trying to figure out how to make a one-off part $20 cheaper? For that matter, how many days do you want to spend designing something, when you could otherwise be making money doing what you're really good at?

Click to expand...

Eggzactly !

As I tell people who are trying to skimp on things (For example, last job was a tractor back blade).

"If you take all this time and energy building it, and get it .001" too thin,
and it folds up, you've wasted all that time, effort & money.
However, if it's 1/2" too thick, and never fails, you've only overpaid for the 1/2"
material, that is your insurance payment"