What's new
What's new

New Pet Peeve- CAD drawing not in Proportion.

Either way, things are not handled will in CAD and/or CAM software. .


Actually, I think they are - at least for the time being - handled well in CAM software.
Meaning: a .5000 hole will be programmed to be a .5000 "feature" based on the CAD file.

If however that .5000 "feature" is a shaft with +.0002/+.0005 tolerance, or a bore with a -.0002/-.0005 tolerance, guess what's gonna happen!!!!


Basically what it boils down to is this:
1: You give me no drawing but a solid model that does not reflect middle, you will get a part centered around the model dimensions you gave me:
If it doesn't fit: It's on you and you pay for it.

2: You give me a drawing with tolerances, and you give me a model that does not reflect middle, then I will re-draw and program to the middle:
It's all on me, but you will pay for me to re-draw and program around it.

3: You give me either a drawing or a model or both that reflects middle:
It's all on me, PERIOD!!!

Engineers, inspectors or purchasing agents need not applied!

Please, take pick your choice between 1, 2 or 3.
 
Were i use to work the cad had about 5-10 years before been done with extensive txt edited dimensions, the costs and fuck ups they caused ran into the tens if not hundreds of thousands over the years post that.

I would with out flinching ditch any customer that still did it, its about the worst offence a drafts man can make in the digital age!
 
Actually, I think they are - at least for the time being - handled well in CAM software.
Meaning: a .5000 hole will be programmed to be a .5000 "feature" based on the CAD file.

If however that .5000 "feature" is a shaft with +.0002/+.0005 tolerance, or a bore with a -.0002/-.0005 tolerance, guess what's gonna happen!!!!


Basically what it boils down to is this:
1: You give me no drawing but a solid model that does not reflect middle, you will get a part centered around the model dimensions you gave me:
If it doesn't fit: It's on you and you pay for it.

2: You give me a drawing with tolerances, and you give me a model that does not reflect middle, then I will re-draw and program to the middle:
It's all on me, but you will pay for me to re-draw and program around it.

3: You give me either a drawing or a model or both that reflects middle:
It's all on me, PERIOD!!!

Engineers, inspectors or purchasing agents need not applied!

Please, take pick your choice between 1, 2 or 3.

Wait a minute, are you saying you are going to 're-draw' the model to nominal? :skep: Um, why?? I think having a print and model (as long as customer says xx is the bible- usually print form what I have seen) is the best of both worlds. If the model has a few features not at nominal, make an adjustment in your program, a simple +/- on 'stock to leave' or 'blank size' will most likely take care of it.
Notice I said a few, if you do a quick check and notice alot of features are not nominal, I might make a phone call...
 
...if you do a quick check and notice alot of features are not nominal, I might make a phone call...


Sorry Mike, I picked much of your response out and using whats left in a totally different manner in which was meant.

BUT...this adds to the Pet Peeve.

Simple job, simple tolerances its a straight forward no-brainer basic machine work job. Face part, couple steps and some holes...only hitch is the jobs a rush.

End of day, machine is between setups, I can stay and run job tonight and she'll be back to regular job in morning. Sweet gravy job, no backups.

I start laying out job in CAM and notice the steps do not add up to the parts overall length. Holes are drawn in relation to steps...call goes into customer and we go over...not quick and easy to explain as he's "made before and his drawing is right". No worries its not important...just put holes as drawn and don't worry about the overall length.
I go back to laying out part and holes are smack in the middle of the steps...can't be right...I take picture of rendering and send to customers cell phone.

Quick call back from him and says that's not right...No Duh!

But he can't answer now as he's not in shop to check, he'll let me know 1st thing in morning...only he gets in a couple hours after we start.

He calls, we go over...and over and over. Finally get it figured out, we make...but now I lost a good portion of day making that rush job and behind on the job that was supposed to be running. Yes I can bang the customer for "his" screw up...I'd be right to...but

Its a Pet Peeve as this stuff is a waste of time, machine time and causes aggravation where I do not need it.
Yes, dump the customer, charge the customer...but its a double edge sword as they can be very good at times...mostly after all the hassles are worked out and I'm in it for the return work more then the 1Zy 2zy's rush stuff.

As I said..Pet Peeve
Sometimes I just wish I could Just machine parts...
 
Out of proportion drawings......??

You ain't since nothing until you look at some of the old Boeing dash-number drawing. Confusing to say the least when in the extreme case the part barely even resembles the drawing.

Yep. Like when a different dash-number made the part in the opposite hand of what it sorta looked like in the drawing.

One had get really good at keeping track of the ADCNs and DCNs on those parts too. Also loved the dash-number drawings where each dash number had a table of dimensions applied to it.

I remember one where the table of "tangent" points and radii calculated out to make a part something like 10" longer than the drawing (on a part only ~30" long to begin with). When you are calculating "tangencies" and radii with a slide-rule versus a computer or calculator you can end up with some big differences.
 
. Also loved the dash-number drawings where each dash number had a table of dimensions applied to it.

.


Boeing ain't got a lock on that one!
Have 2 Sikorsky details where there aren't a single numerical dimension by the "doodle" other than:
"Fitting end as-per AMS53649-16, except see table."
 
Boeing ain't got a lock on that one!
Have 2 Sikorsky details where there aren't a single numerical dimension by the "doodle" other than:
"Fitting end as-per AMS53649-16, except see table."



Thanks, that made my day.

I forgot all about those As per MIL-F-XYZ-FU prints. I got roped into doing that twice...the next time I handed it back to customer telling him you pull every spec and enter onto print. It's more work finding the specs and making an in-house print to go by then making the actual part. Way too easy to make a mistake. I have No Head for that.
I'll take my non-proportional drawings over that nonsense. No wonder palnes and military stuff is so expensive.
 
.....the next time I handed it back to customer telling him you pull every spec and enter onto print. It's more work finding the specs and making an in-house print to go by then making the actual part.

I spent many hours going through a local university's collection of MIL documents back in the 70s and 80s looking up specs.

No wonder palnes and military stuff is so expensive.

I remember one job making cable tensioning cylinders for Army mobile bridges where the seal sets were a bit over $500 per set. For exactly the same seals, but non-MIL spec parts, was less than $30.
 
I was told, "make it to the print dimensions. If it doesn't work, its not your problem". This all goes back to engineering. Drafting is not qualified to add tolerances. I have said this before and I will say it again - the specifications for a part or assembly are TOTALLY the responsibility of the engineer. Single sided tolerances should NEVER be used. A tolerance analysis should provide a reasonable value. Example: X+.010-.000 is totally meaningless. If the part can function properly at X+.010, then why not set the value to X=-.005? That's the way I would make the part. Second peeve X-.02-.05. The has shown up numerous times on metric drawings. I don't know if this someone's wet dream or if there is a standard, but it makes no sence at all.

I spent my career as a design engineer. I have worked with draftsmen and I have made drawings complete on the board and on AUTOCAD.

Now I make parts from other peoples nightmares. I have made drawings from other peoples drawings so that I could figure out what they were trying to. And somehow we made to the moon and back.:eek:

Tom
 
Single sided tolerances should NEVER be used. A tolerance analysis should provide a reasonable value. Example: X+.010-.000 is totally meaningless. If the part can function properly at X+.010, then why not set the value to X=-.005? That's the way I would make the part.

Why not? Single sided tolerances don't prevent choosing functional tolerances.


Second peeve X-.02-.05. The has shown up numerous times on metric drawings. I don't know if this someone's wet dream or if there is a standard, but it makes no sence at all.

Yes, they are from ISO 286. It makes a lot of sense. When the letter/number code is given, it shows the intended fit when the part goes into assembly. As the number gets smaller, the tolerance range gets smaller. As the letter moves towards the end of the alphabet, the fit gets tighter. F or f is a loose fit, S or s is a tight fit. The info should be in the tech section of every metric bearing catalog.
 
Why not? Single sided tolerances don't prevent choosing functional tolerances.




Yes, they are from ISO 286. It makes a lot of sense. When the letter/number code is given, it shows the intended fit when the part goes into assembly. As the number gets smaller, the tolerance range gets smaller. As the letter moves towards the end of the alphabet, the fit gets tighter. F or f is a loose fit, S or s is a tight fit. The info should be in the tech section of every metric bearing catalog.

And where there is no letter associated with the dimension? Then what?

Tom
 
And where there is no letter associated with the dimension? Then what?

Tom

I'm not sure what you mean. To call the tolerance by ISO code, it needs both the letter and number to define the tolerance. Just the number would only give you the tolerance grade. That would be a drawing error.
 
He never said it was!! You did.

Saying that something is an ISO standard does not make it right-as deemed by God (let alone good).

R

True, I brought up ISO 286. Plus/Plus and Minus/Minus tolerances are part of the standard, and the standard works well when it is used appropriately. My experience is that it does what a standard is supposed to do, and gives consistent fit tolerances when different designers are making different parts that go together.
 
I obviously do not know ISO 286, I looked at it and came away confused. That's neither here or there.

I was taught that a dimension consisted of a nominal or "perfect" value with acceptable limits. I was also taught as a machinist to aim for the "perfect" dimension as closely as practical. In the first case of the off center tolerance, that means that I would set the machine to the "perfect" value. Statistically, that would result in 50% rejects, so I set the nominal value to the center of the tolerance band. This throws work on the setup person and places the burden of deciding what the nominal dimension should be on the shop, not the engineer. An opportunity for errors to be made.

The second case is a mystery of how its supposed to work, especially when bastardized by someone that doesn't know how to use it as was in the case presented.

I guess I am just getting too old. Time to hang up the micrometers.

Tom
 








 
Back
Top