Site Notices
8/19/2014 4:57:02 PM
Author
Message
Happygilmoresar
Offline
Posts: 1
Feedback: 0% (0)
Posted: 4/16/2013 5:52:11 PM
Hey guys,
New to the site here. I am going into production of upper and lower receivers for the first time. So far I have been unable to find an acceptable blueprint for the upper receivers. Mostly what I need is tolerancing on the critical sizes for universal compatibility of all mating components. If anyone has a link to help me out that would be great. Thanx.

Dave
ar15parts@aol.com
headmonkey
Offline
Posts: 505
Feedback: 0% (0)
Link To This Post
Posted: 4/17/2013 12:13:52 AM
[Last Edit: 4/17/2013 1:42:47 AM by headmonkey]
Originally Posted By Happygilmoresar:
Hey guys,
New to the site here. I am going into production of upper and lower receivers for the first time. So far I have been unable to find an acceptable blueprint for the upper receivers. Mostly what I need is tolerancing on the critical sizes for universal compatibility of all mating components. If anyone has a link to help me out that would be great. Thanx.

Dave
ar15parts@aol.com


You need a probe.

Budget priced ebay probe

If you say you don't need a probe, then please put me on your 'do not sell product to' list.

Okay ... I'll explain.

Anything you attempt to manufacture, will not be what you actually end up manufacturing, except, X percent of the time.

While you will design the part dimensions to be exact, the machine, the cutting bit, the time of day, the operator, whether or not
the operator got laid last night, the material, the temperature, the position of the sun in the sky, whether or not Obama is
golfing or drinking that day, etc. will all conspire against you, and the part you attempted to produce with exact dimensions,
will not be the part the machine cranks out and ends up in your hand.

This is why manufacturers have quality control processes in place. You will need to remove X number of parts from every batch of 100 or
so and probe them to make sure that your machines are actually producing the components within the allowable tolerances.

In addition, you'll be sending these parts out for coating / anodizing etc. You'll have to measure X number of each batch that returns
from the coating subcontractor, or department if you do it yourself.

You need to monitor the dimensions of the actual parts produced on a continuous basis, unless you plan on sticking your head in the sand,
pretending they are all perfect because "they look real nice," and letting your customers via feedback tell you when your processes have
drifted out of spec.

Screw 1 customer because you failed to perform QC and they'll tell the entire internet you suck, and it'll be on the internet for forever.

Make 10 customers really happy, and consider yourself lucky when 1 of them tells the entire internet your part was acceptable.

As an extreme example, I used to work for a well known soft contact lense manufacturing plant. They would plan a production run to center
on whatever prescriptions they needed according to inventory, but in each production run, X numbers of lenses would hit the target, and Y
number of lenses would land entirely in other prescriptions. They then used a special laser machine and measured each and every single
contact lense after hydration, and recorded its parameters before placing it in inventory, because they are not able to control what they
actually produced. It's not nearly as bad with AR components, or parts machined from aluminum in general, but you get the idea.

This is standard manufacturing industry practice.

Statistical Process Control

I would say 'Good Luck', except, I just explained how to remove luck from the process.

Making a few parts, does not equal, operating a manufacturing business successfully.

ps. This was focused on customer satisfaction. I didn't mention what could happen to you financially or even
criminally if your components explode while a customer is using them, you failed to perform quality control, and
cannot provide data to prove that you produced the parts properly.

I'll let your imagination run wild with the consequences.

While you may not actually be at fault, without data to show you are not at fault, you may very well end up
learning the definition of the term 'scapegoat' the hard way. All it takes is 1 accident, and an attorney who
has more time and money than you do, and you'll be closing shop, and flushing all your hard work down
the toilet.

pps. In addition to a probe, you may want to plan on using X-ray or ultrasound to verify that there are no
voids internal to your materials that would create a weakness. Probably more expensive than $200, but
without performing this quality control check, your pants will still be down around your ankles.

Then there will be surface hardness testing, color testing, finish abrasion / corrosion resistance testing,
dye penetrant inspection, etc, etc, etc.

Looking forward to reading on your website the great lengths your company goes through to insure the
highest quality components.

GTBJW
Helluva Engineer
NRA
Offline
Posts: 299
Feedback: 100% (7)
Link To This Post
Posted: 4/17/2013 3:02:43 AM
This thread has potential.
ΜΟΛΩΝ ΛΑΒΕ
GHPorter
Crufflers Rule!
Military
Offline
Posts: 11930
Feedback: 100% (4)
Link To This Post
Posted: 4/17/2013 6:47:16 AM
QC is a make-or-break thing for a manufacturer, both for public opinion and bottom line.

As noted earlier, ineffective QC will easily earn you a reputation as a hack manufacturer, and that is a very deep hole that most can't dig themselves out of. But even with the best of intentions and a serious plan to make your customers happy, it is very expensive to take back parts and replace them. The more important and complex the part that goes out out of spec, the more expensive this will be. It is essentially an industry standard for the manufacturer to cover shipping in both directions for exchanging bad parts, and that can cut into the bottom line very quickly. It also takes time and manpower away from selling and shipping parts you're making money off of, and the worst thing you can do is to put off dealing with replacements in favor of sending out new orders, because replacement parts are essentially parts that are shipped out VERY late.

You will also need to examine the returned bad part to find out where the part went wrong so you can fix your processes to avoid repeating that particular problem. This is usually difficult and time consuming, and bean counters consider it a drain on productivity because it is a preventive process (and you can't explain the payback of any preventive process to any bean counter).

Everybody hates the QC department, but any attempt to do without that department is essentially a plan for the entire enterprise to fail. Whether you make uppers, or trigger guards or sling adjuster loops, if you don't put a lot of time and effort into making sure what you send otu is "right" before it goes out, you may as well just not start in the first place. It is that simple.
"--you can't conquer a free man; the most you can do is kill him."
Heinlein
Happygilmoresar
Offline
Posts: 2
Feedback: 0% (0)
Link To This Post
Posted: 4/17/2013 1:18:50 PM
thanks for the cold shower guys. I simply asked for for a blueprint of the upper receiver with proper specs that I could use. Any help would be greatly appreciated.

glycerine
Offline
Posts: 114
Feedback: 100% (4)
Link To This Post
Posted: 4/17/2013 1:27:03 PM
Using a probe will allow you to basically generate your own "blueprints" from an existing part, right?
Happygilmoresar
Offline
Posts: 3
Feedback: 0% (0)
Link To This Post
Posted: 4/17/2013 1:50:16 PM
yeah it is. Only a complete hobby amateur would use a $200 Chinese probe and copy a part. Then to use that information to make 100's or 1000's of parts off of that data.
glycerine
Offline
Posts: 115
Feedback: 100% (4)
Link To This Post
Posted: 4/17/2013 2:23:49 PM
I'm sure you could find a more expensive one...
headmonkey
Offline
Posts: 506
Feedback: 0% (0)
Link To This Post
Posted: 4/17/2013 2:41:43 PM
[Last Edit: 4/17/2013 3:04:15 PM by headmonkey]
Originally Posted By Happygilmoresar:
yeah it is. Only a complete hobby amateur would use a $200 Chinese probe and copy a part. Then to use that information to make 100's or 1000's of parts off of that data.


Correction ... Only a complete hobby amateur would read this thread, and completely ignore the discussion of quality control, which is the reason you need a probe, and then post a smart ass comment about using a probe once to make copies without QC.

I'll try to explain it again, in simple easy steps for you.

You need a probe for quality control, get one. You can spend $200 on a not-so-shitty probe for initial setup, and use statistical techniques such as regression analysis to correct for any definiciencies in the probe itself, or you can spend $2,000 or $20,000 ( it's your money ) on a "high quality probe" and you'll still need to use the same statistical tests to insure the probe is functioning correctly. It's your business, buy whatever probe you want, point is, you can't do QC without one, so you may as well get it now, since you plan on being a big time manufacturer, and use it for design as well.

Data is data. You can get it from a milspec data sheet, or you can enter it with a probe. I'm sorry I didn't link to all the software you will need with the GCODE copied from DPMS or some other manufacturer.

Do your own homework Johnny ... no blaming the dog when your non-QC'd parts kill someone.

And before you spend $200 or $2,000 or $20,000 on a probe, do everyone a favor and spend $2 on a book about manufacturing processes, becaue ATM all you appear to have is a dream.

A smart business person would get the $200 probe, because he realizes it's only 1 of 1,000 pieces of the puzzle, and after using math to validate the data collected, if he discovers its inadequate for the task, then move up to the $2,000 level.

Sorry if operating a manufacturing business isn't as simple as chucking a block of aluminum into a vise, pushing a button, and incrementing the inventory counter on your website.

Perhaps the 'cold shower' was really a 'wake up call' that perhaps there are issues you have not considered.

From what I see, everyone, myself included, is trying to help you.

As with all free advice, take it or leave it, and I wish you all the best on your journey of discovery. If you don't give up, it could be quite the accomplishment one day.

krrtlr1
Offline
Posts: 54
Feedback: 100% (1)
Link To This Post
Posted: 4/17/2013 3:16:59 PM
Originally Posted By GHPorter:
QC is a make-or-break thing for a manufacturer, both for public opinion and bottom line.

As noted earlier, ineffective QC will easily earn you a reputation as a hack manufacturer, and that is a very deep hole that most can't dig themselves out of. But even with the best of intentions and a serious plan to make your customers happy, it is very expensive to take back parts and replace them. The more important and complex the part that goes out out of spec, the more expensive this will be. It is essentially an industry standard for the manufacturer to cover shipping in both directions for exchanging bad parts, and that can cut into the bottom line very quickly. It also takes time and manpower away from selling and shipping parts you're making money off of, and the worst thing you can do is to put off dealing with replacements in favor of sending out new orders, because replacement parts are essentially parts that are shipped out VERY late.

You will also need to examine the returned bad part to find out where the part went wrong so you can fix your processes to avoid repeating that particular problem. This is usually difficult and time consuming, and bean counters consider it a drain on productivity because it is a preventive process (and you can't explain the payback of any preventive process to any bean counter).

Everybody hates the QC department, but any attempt to do without that department is essentially a plan for the entire enterprise to fail. Whether you make uppers, or trigger guards or sling adjuster loops, if you don't put a lot of time and effort into making sure what you send otu is "right" before it goes out, you may as well just not start in the first place. It is that simple.


QC department doesn't create quality. It only validates it...... or doesn't.

Dr69er
Offline
Posts: 1203
Feedback: 100% (5)
Link To This Post
Posted: 4/17/2013 4:55:36 PM
[Last Edit: 4/18/2013 9:17:00 AM by Dr69er]
You can also use a laser with a DRO on your machine, should work pretty good as well...

I have 8 pages worth of blueprints for a std. 7075 T6 flat top AR Upper somewhere in the shop...

Contact me via email if your Interested in purchasing the prints...
iNeXile556
Deo vindice
Military
Offline
Posts: 3356
Feedback: 100% (22)
Link To This Post
Posted: 4/17/2013 5:25:43 PM
[Last Edit: 4/17/2013 5:51:39 PM by iNeXile556]
So let me get this straight, your a FFL07/03 with a full CNC machine shop that doesn't know where to get the specs for the most popular firearm in the United States.
The world is a dangerous place to live; not because of the people who are evil,
but because of the people who don't do anything about it.

Rise And Rise Again Until Lambs Become Lions.
GHPorter
Crufflers Rule!
Military
Offline
Posts: 11931
Feedback: 100% (4)
Link To This Post
Posted: 4/18/2013 6:42:26 AM
Originally Posted By krrtlr1:
QC department doesn't create quality. It only validates it...... or doesn't.

Absolutely true. But if your QC department doesn't keep an eye on your product, your quality can turn out to be hit or miss, and process problems can turn into business killers. Being "dedicated to quality" won't help if you don't monitor that quality.
"--you can't conquer a free man; the most you can do is kill him."
Heinlein