So what's enough tolerance for "YOU"!

[modamag]

A few of you probably know that I collect flashlight for just the pure physical property of the light. Probably a few groove here and few slots there.

As I slowly migrate from a consumer -> designer -> fabricator. I found one of the most intriging aspect of how lights are produced.

There are bassically two schools of thought.

1. Keep everything loose and design for production (+0/-0.005") eg. MagLites
2. Keep them tolerance as tight as possible (+0/-0.001") eg Larry stuff.

Now both of these school of thoughts requires alot of time in either way.
=== The loose thinking ...
For these guys alot of design and modeling have to be done up front to ensure the "worst" case scenario still works properly. They probably spend 2-3x more for the design than most case.

But when it comes to manufacturing, with 5 thous tolerance they can just crank up the speed, bar stock feeder and run 24/7 ... well almost :nana:

=== The mens in tight gang ...
These guys have little need for modeling since on paper everything fits just right! Oring groove are a perfect 18% fit, walls are 0.010" thick, while thread pitch gap are just perfect that even water moisture don't even seep thru there, or do they. :devil:

When it comes down to making it, they whip out their pin gauge, thread pitch gauge, keep a thermometer near by and spend hours on ends measuring and test fitting to ensure that the right "feel" is there.

.
..
...
....
.....

So what is this all about? Well I would just like to know which groups of tolerance you are in.

Mens in tights ... 0-0.001" tolerance
Mens in reds ... +/-0.005" tolerance
I own a ruller ... 1/16" tolerance
That looks right ... minimal use of measuring devices.

I originally was in the red (chinese machine), then went to the tights, but recently found a way to appreciate both schools of thoughts.

 

[DonShock]

On the purchasing side, I'm "Mens in reds ... +/-0.005" tolerance" for ease of usefor battery swaps, etc. Although I do love my Draco and I can't imagine it would have been possible to build it that loose. But as far as my own mods, I'm still at the "I own a ruller ... 1/16" tolerance" stage right now. My method of shortening a Mag involves a plastic miter box, a hand hacksaw, and JB-Weld.

 

[PhotonFanatic]

0.01mm tolerance for me.

Well, at least that's what my DRO can distinguish. :lolsign:

 

[will]

I always like the answer 'It Depends'

If you are going to press fit something together - you are living in the world of .0001 inch

If you want 2 pieces that will slide together like a rod inside a sleeve. Then you are working with a minus tolerance on the inside part and a plus tolerance on the outer part.

Hole locations seemed to be mostly + or - .005


Tolerances on the outer part of a flashlight can be pretty loose. The outside does not have to fit into something else, if you will - it is for looks.

Many complex parts that are machined have different tolerances for different areas of the part. It depends on what additional parts are going to be assembled or fitted on the different areas..

You really have to look at the application and what parts are going to be assembled, which parts have to move, which parts will not move. Anything that has to move should move freely, but not be sloppy. That is why some mass manufactured items are prototyped first, to see how it fits and how it works.

Generally speaking - the closer the tolerance, the more you have to check, the greater the reject rate - this translates to greater cost.
Like I said ' It Depends'

 

[gadget_lover]

I like Them tight. That's what makes them different from the Walmart imports.

I have a nice pretty chrome plated MXDL. It has threads at the head that are so loose that the head rocks back and forth of it's not screwed all the way down. If it's not tightened down the head WILL fall off eventually.

OTOH, I have an ARC that has nice tightly matched threads and closely machined tolerances. It feels nice and smooth.

My edc has very tight tolerances throughout. It has no o-rings, yet there's not a lot of room for water to penetrate the threads in a quick dunk, rain, etc. I feel good about that light simply because every part of it is EXACTLY the size that I set out to make it.

Daniel

 

[KC2IXE]

AAARRRGGGHHHH - OK, this is a perpetual argument in the Model Eng and Live Steam fields. You'll often hear "if the tolerances are too tight, it won't work"

I have to point out - there is a difference between tolerance and FITS. I'f I'm fitting a 1/8 pin into a 1/4 hole, it's a VERY loose fit. However, you could spec that 1/8" pin +-.0001, and the 1/4 hole +-.0001. It's a very tight tolerance but a very loose fit

FITS must be loose enough to not jam up, and tight enough to do what is intended. Tolerances must be correct to get the right fit. If you are NOT doing interchangable parts, tolerances can be very loose, so long as you make the mating parts to fit

Fixed the 2nd pin from 1/4 to 1/8 - typo

 

[will]

Like I said earlier - it depends on what you are making..

 

[will]

.

 

[petersmith6]

8thenths of a hundreth of a micron...i use to work in a calibration lab.i calibrated master gauge blocks.the fun bit is if the surfaces are sooo flat that when wrung togeter they will stick as you have squeezed all the air out of the joint.+-.0001 of an inch is huge

 

[cy]

another measurement nut here...

my bench includes several calibration grade tools.

 

[gadget_lover]

One of the things I've noticed is that I can feel the difference between two parts that are screwed together if one is 1/1000 smaller than the other. It's not a big difference, but when you handle it the seams are noticeable. The trick, of course, is to make each part significantly different sizes so there is no need to match the dimensions.

Daniel

 

[modamag]

I think KC2IXE capture what I wanted to say the most.

PhotonFanatic: Fred you know that's a curse when your DRO goes in the tenths right! Only in my dreams do I own a DRO that does +/- 0.01
What you don't know won't hurt you

petersmith6: "hundreth of a micron" eh, That sound like my kind of work 45nm. Too tight of a tolerance for me to have fun.

 

[McGizmo]

I would like to consider myself a realist as well as a pragmatist. When I design something, first and foremost, I want it to work and I don't want exceptions that need TLC. I have been influenced by what I have heard called "marine tolerances" in the marine industry and these are typically "loose" in anticipation and allowance for external "stuff" to be part of the assembly in addition to the original components. External stuff may be grease and lubes as well as some salt build up. In a world where dust bunnies and belly button lint run free, one must allow for their presence or suffer the consequences.

I have components made by different shops and some work with well seasoned screw machines where others have the latest and greatest CNC equipment. I don't pretend to be an engineer but I do know I need to design against interference fits of components that need to move freely. I have been moving to specifying max and min dimensions on critical mating geometries. I figure if I say specify a .002" gap beween a Min bore and Max OD that the likelihood of an interference fit is slim and if the shops making the parts are working within a +/-.002 range that the typical paring will be just fine.

It also seems a good idea to try to design with the intent of reducing the number and dependence of tolerances being stacked.
I consider myself somewhere between tight and red as defined here.

 

[PEU]

In my case I'm lucky enough to be able to chat with the CNC machinist before, while and after the machining process of every piece I make.

Take for example a sealing oring that also offers some resistance to the turning of a twisty body, spec sheets weren't made for flashlights, so the best solution I found is to be there when the oring placeholder is being done with the oring and feel if the pressure is right, if not, ask machinist to lose or add some 0.01mm to the mix and do a new piece.

Yeah, it takes time, but we all know, its far easier to remove material than to put it back

I tend to stay on the tight side of the tolerances, not the extreme, but tight.


Pablo

 

[b2eze]

"In a world where dust bunnies and belly button lint run free, one must allow for their presence or suffer the consequences."

MCGIZ! You must have been Lord Byrum in the last life! Pure poetry! I love it!

 

[RCatR]

I can pull +/- .0002" on the mill and +/- .0005 on the lathe; but in flashlight applications I like to go with the +/- .005(ish); it seems to make the threads work ALOT better

 

[highorder]

on most work, I think that +/-.002 is a decent compromise.

If it has to be right on, +/- .0005

 

[modamag]

b2eze: Didn't you know, McGizmo live in paradise. Some of those poetic stuff must sink in somehow.

RCatR: I would think it's the opposite.
I have harder time holding tight tolerance on my mill than lathe? WTH?
How do you hold 2/10th on a mill? Or is that your encoder position accuracy (repeatability)?

inquiring minds wanna know ...

I can pull +/- .0002" on the mill and +/- .0005 on the lathe; but in flashlight applications I like to go with the +/- .005(ish); it seems to make the threads work ALOT better