What's new
What's new

Inspecting Rounded and Drafted Edges

S3D_Adam

Plastic
Joined
Nov 6, 2015
Location
Woodinville, WA
Does anyone have any tips for measuring dimensions on drafted and rounded edges? I can never seem to find a consistent edge with an OMM, and with a CMM, it is difficult to measure points at the correct z-height. See the picture below for the type of dimension I'm talking about.

DRAFTED_AND_ROUNDED.jpg
 
Roll pin. I did it a lot as a tool and die maker 20 years ago. You need to know the diameter of the pin and angle of your reference surfaces, then trig it out. Not sure how to measure your radius except for a radius gage.
 
When I have something like that I use a linear scan on the profile of the draft so I have a "linear" pointcloud. Then I use PCDMIS to give me some geometry with a least square method. Then you can pop it in AutoCad and play with it to your heart's content.
 
When I have something like that I use a linear scan on the profile of the draft so I have a "linear" pointcloud. Then I use PCDMIS to give me some geometry with a least square method. Then you can pop it in AutoCad and play with it to your heart's content.

lol I was going to make some kind of similar smart assed answer but I think one's enough for a thread.


@AWY: that's the two options I'd use with the tools I have on-hand - Bridge CMM or manual CMM/scanner or failing that, the tried and true pins and mics with a little math as described already.
 
A manual cmm or a cnc/dnc cmm.
If a cnc style this should not be a problem so I'm thinking manual. On a cnc/dnc you just have to program it correctly.
On a manual you can blow the Z height by a ways but let the computer calculate the two lines.
Rads are always a matter of how you want to check, 3 point, 20 point best fit, max deviation from true to form. Add in tangent to parts sides.

If a cast part then cmm single points might not reflect the actual surface so you can't do two point lines in space. Measuring over rolls have the same problem on cast.
Least squares fit also can and will screw up here although not very often.
No clues here as to size, tolerance or surface finish on the walls and rad.
Bob
 
I've never used the roll pin idea before. Am I correct in thinking that I can use the roll pins to measure the draft angle, and not the width of the feature as specified in the picture I originally attached? Or can I use roll pins to inspect the width too?
 
Do you have any cad software, and/or do you have this part electronically? IF yes to both, draw a couple circles at the tangent of side and top, measure across pins in cad, and mimic the process using an angle plate and surface plate, height gage, pins.
 
Yup. That's the concept IMO, side to side measurement and a bunch of trig. Make sure you use enough decimal places to afford some rounding error. Think about any practicalities of measuring tool capacity vs. pin size.
 
Does anyone have any tips for measuring dimensions on drafted and rounded edges? I can never seem to find a consistent edge with an OMM, and with a CMM, it is difficult to measure points at the correct z-height. See the picture below for the type of dimension I'm talking about.

View attachment 174802

I'm going to end up being known as the guy that asks questions more than answers them.

Have looked at your sketch a couple of times and am not 100% sure of what you are wanting. Is it what I have on my sketch?

Distance.jpg

If it is then my answer would vary depending on the quantity to be measured plus size and the tolerances and thus accuracy required.

Is it a "solid" object or sheet metal?

Just because someone has a CMM doesn't mean everything has to be measured using it.
 
We inspect a lot of small plastic injection molded parts. Many times there will be a small screw boss protruding from the "solid" part. The drawing will have a width dimension to the tangent of either the top fillet tangents or the bottom fillet tangents that will state a dimension and then either +draft or -draft.

Gordon's sketch shows a width that I may need if I use the roll pin method. I've never use the roll pin method, I've never seen anyone use the roll pin method, and I haven't been able to find a visualization on the internet anywhere, so I'm not even sure if I need that dimension.

Perhaps I just need to purchase a copy of ASME Y14.8. I'm not even sure if the engineers are dimensioning these types of features properly.
 
We inspect a lot of small plastic injection molded parts. Many times there will be a small screw boss protruding from the "solid" part. The drawing will have a width dimension to the tangent of either the top fillet tangents or the bottom fillet tangents that will state a dimension and then either +draft or -draft.

Sounds like it's a round boss. If so forget the roll pin method.

Gene
 
We inspect a lot of small plastic injection molded parts. Many times there will be a small screw boss protruding from the "solid" part. The drawing will have a width dimension to the tangent of either the top fillet tangents or the bottom fillet tangents that will state a dimension and then either +draft or -draft.

Gordon's sketch shows a width that I may need if I use the roll pin method. I've never use the roll pin method, I've never seen anyone use the roll pin method, and I haven't been able to find a visualization on the internet anywhere, so I'm not even sure if I need that dimension.

Perhaps I just need to purchase a copy of ASME Y14.8. I'm not even sure if the engineers are dimensioning these types of features properly.

I'm understanding less and less. I am getting now that it is a solid part but now I also have learned that they are plastic parts from injection molds.

1. What size are the parts?
2. How many are produced by the mold at one time? I don't mean production quantity.
3. Has the injection mold been made by the same company/shop as that making the items?
4. Injection molded item are usually mass produced. Is this the case here and is what is being done a FAI?
5. What are typical tolerances?
6. It might be relevant to know what type of plastic.

Without answers to these basic questions then any answers given would be guesswork and speculation. A "typical" drawing (as there would seem to be several part numbers) would also be an immense help.

BTW the "circles" in my sketch weren't roller pins. They were radii to show the specified distance of X.

I'm only ½ an hour from LEGO where I live so I know quite a few LEGO injection mold designers and builders and it doesn't get much better than that.
 
I work in a metrology lab that was built to do short runs and one-off parts. My primary function is to help engineers solve issues quickly. Our parts are manufactured overseas, so I never know very much about the parts I'm inspecting. I am usually just given 2 or 3 to find out why something isn't fitting or if the engineer doesn't believe a measurement that the vendor made.

Guesswork and speculation is exactly what I'm after. I'm not looking for one solution to perfectly match one problem. Instead, I want to know what options are available so that I can be better prepared the next time someone brings me a molded plastic part to inspect.

I'm afraid that the only question I can really answer is regarding size and tolerance. Parts typically range anywhere from about 5 to 300 mm in length, but the typical size and tolerance on a dimension like the one in the image in my original post would be 3+/-0.15 mm.

Sorry I can't give more information than that. As it stands, I think the only option I have to go on is Rewt's suggestion (post #3).
 
I work in a metrology lab that was built to do short runs and one-off parts. My primary function is to help engineers solve issues quickly. Our parts are manufactured overseas, so I never know very much about the parts I'm inspecting. I am usually just given 2 or 3 to find out why something isn't fitting or if the engineer doesn't believe a measurement that the vendor made.

Guesswork and speculation is exactly what I'm after. I'm not looking for one solution to perfectly match one problem. Instead, I want to know what options are available so that I can be better prepared the next time someone brings me a molded plastic part to inspect.

I'm afraid that the only question I can really answer is regarding size and tolerance. Parts typically range anywhere from about 5 to 300 mm in length, but the typical size and tolerance on a dimension like the one in the image in my original post would be 3+/-0.15 mm.

Sorry I can't give more information than that. As it stands, I think the only option I have to go on is Rewt's suggestion (post #3).

Open your cad, find dimension A - from the tangent of top radius to tapered wall section, find dimension B - from bottom flange of the part to the point you just found on A. Now, the 'guesswork' part. Use a scale, or block setup of some sort, now measure up from flange and set your calipers there and measure. Ya, it is a little (a lot?) hillbilly, but you should get a ballpark to see if you are in the range. If the measurement checks close to the print, you can either investigate further using a waaaayy better method, or start looking elsewhere on the part for interference...

edit: What I have posted wont get you very close to +/-.15mm (.005"), BUT it should tell you if something is way off... without seeing function and the actual part all of this i guesswork...
 
Two ways that I would do it:

CMM- Establish Z plane on the part and zero to it. Come off of Z what ever the required distance is for the "gage point". Lock Z axis and take the required hits of the diameter.

Optical Comparator- Basically the same as above.

On the right Comparator, it would take less then a minute and would be very repeatable.
 
I work in a metrology lab that was built to do short runs and one-off parts. My primary function is to help engineers solve issues quickly. Our parts are manufactured overseas, so I never know very much about the parts I'm inspecting. I am usually just given 2 or 3 to find out why something isn't fitting or if the engineer doesn't believe a measurement that the vendor made.

Guesswork and speculation is exactly what I'm after. I'm not looking for one solution to perfectly match one problem. Instead, I want to know what options are available so that I can be better prepared the next time someone brings me a molded plastic part to inspect.

I'm afraid that the only question I can really answer is regarding size and tolerance. Parts typically range anywhere from about 5 to 300 mm in length, but the typical size and tolerance on a dimension like the one in the image in my original post would be 3+/-0.15 mm.

Sorry I can't give more information than that. As it stands, I think the only option I have to go on is Rewt's suggestion (post #3).

Not trying to get you unemployed but it does surprise me that your customers doesn't get a 3D print of the part based on the drawing before a mold even gets made. This can also be done before anything gets started overseas.

If something didn't appear to be OK with the 3D printed item then time and money would be saved.

The 3D printed item can then be compared to what the mold produces. If it doesn't then your customer can send it to you.

Considering what a good precision mold costs it's certainly the way I'd go. Just for the record the distance tolerance between finished LEGO brick knobs is 8mm ±0.02mm / ±0.0008"
 








 
Back
Top