I would encourage using a symmetry axis and dimensioning over it. Definitly using the a standing dimension and dividing it is an Okay way but using an equal constraint and one dimension will look cleaner in detailed scetches. I like your information density in your tutorial. Keep up the good work.
Great video, from the subject, information, and process perspectives. Almost takes me back to a time when you did a lot of firearm related work on your channel. Thanks again!
Thank you very much! i was looking for a way to draw and so confused just based on the blueprint/dimensions google provided... THIS IS AN AWESOME WAY TO EXPLAIN IT KEEP IT UP!
John, I have been working on parts for my pcc, I have been struggling with getting the model for the rail correct. Your video was extremely helpful! I see now what I have been missing. Thank you!!!!
There's a new standard now called the NATO Accessory Rail STANAG 4694, very similar to the MIL-STD-1913 with a couple tweaks for how accessories interface, and a big difference in that it's defined in metric. Food for thought!
Wow. I really am having trouble with drawing in 360 after mastering solidworks over the last 20 years. It's a real challenge making the transition between platforms. I can create this in under a minute and would have gone with a center/mirror line feature as a driving start. Nice work. Thanks for sharing your knowledge.
you, my brother from across state borders, you have just earned yourself a new subscriber. email me one day when you get a chance. would love to learn more about cnc.
I don't use Fusion360 (one of the last holdouts!) so I've been skipping the Fusion Friday videos lately but this one caught my eye. You sure made it look easy. There should be an online library of crowd sourced common parts like the Picatinny rail.
I did all the steps, and I have more space on the bottom of that hexagon shape i created than I do on the top. I followed the steps carefully, and it didn't work for me.
Ahhhh Guns... Thanks, I was just about to ask Google what she thought. I get a pass on not knowing that one - I'm Canadian, we prefer Beer, buying 50 year old clunker submarines from other countries and relying on our big brother "downstairs" to take on our fights - and we appreciate it... ;)
At 5:00, you can instead use the "equal" constraint on the two small edge break lines. That original drawing looks a lot more complicated than it needs to be. Is there any reasoning from a CNC perspective to make the dimensions like that?
Well its an old drawing they tend to be more complicated than what they should be. but it contains all the tolerances needed as well. i work in a really old factory its funny to read 50 year old drawing compared to 1 year old drawings.
I would have dimensioned it as a rectangle with four chamfers but maybe tolerances would stack too much. I guess it matters which faces the accessory indexes on.
That drawing isn’t complicated. It’s dimensioned in such a way to utilized GD&T and to ensure that no matter where within the tolerances given that the mating parts will fit correctly.
How do you measure the dimensions after the part is made? For example how do you measure the 0.748 length of the rectangle with calipers? Do you need a robotic probe just to see if it meets spec?
James Patrick It’s true position call outs would pretty much require a CMM to do it. You can do it the old school way with V-blocks, height gauges, gauge pins and such as well. But a CMM would be the best way. If you wanted to measure that .748 you would set 2 pins of equal size (I think .25” in this case) to a center to center distance of .835, set the part upside down to the drawing orientation, then do the math (which I’m not gonna do here) to figure the height distance from the pin center line to the flat (with the .005 flatness call out). That will get you a derived dimension that if met will confirm the .748 dimension is correct.
Excellent demo! Let’s throw a kink in the cog. How would I model 20 MOA cant into the rail over a given distance? For example, my LaRue OBR has a 20 MOA cant starting at the back of the rail and continuing into the hand guard. Do you manually do the trig and input it as a slope or can you do it in the dimension as a calculated value?
Just model it as John did, but to the length you want. Then create a second sketch on the side of the angle you want and extrude it. 20 MOA is just 1/3 of a degree
Do you have any examples of how you start/draw a multi component model? For example I am currently trying to sketch up the Gingery shaper based off of prints in the book in fusion. How do you manage all components, when to do the assemblies/joints, how about fasteners? New components for each one. Sorry total newb here.
Nate Van Bynen new components for each part. You can then export all components as individual components. Then start a new file and bring in each individual part you need to do your assembly and do your joints and such there. Fusion handles assemblies, especially large ones rather poorly. It can be done and have seen injection mold assemblies done in fusion, but it’s rather clunky. The software is clearly built more around the part side of things rather than the assembly side of things.
Occams Sawzall thanks so much for the reply, that makes more sense. I have been tryin to do joints as I go along and after a bit and a few components later I find it just gets to be to much
I guess I am Mr Nit picker this time, but the Picatinny rail was not done - you didn't do the slots!!! Great video as usual, however. Not having Fu360 I just did it straight in Gcode - a suggestion perhaps? :~>
IMHO... something is lost where a standard drawing has dimensioning made from arbitrary reference geometries rather than geometries of the actual part itself. There is a disconnect from the reality of how it would be machined on a manual machine. In my line of work (railroad track) see updated spec drawings for rail profiles and there are dimensional errors on them where transitions between fillets are not tangent to each other at osculating points because some kid fresh out of engineering school, with no background in mechanical drafting, made a drawing however it was easiest to him in CAD.
R.C. Whitehead Been fighting things like that all my life....Engineers driving up the costs of things because they don't realize the ramifications of their incompetence. Try as I might I've tried to train them in some common sense techniques but Every time they take a lunch break it's time to start over. :(
Cad-Cam_Man - I had a mechanical drafting class my freshman year of engineering school. I was perplexed how something simple like orthographic projections and isometrics made absolutely no sense to the females in the class. They had no sense of how something is constructed from 3 reference views. I bet they are now working at large firms with 6 figure salaries.
R.C. Whitehead It’s because that’s the critical locations for the mating parts. And the true position tolerances at maximum material actually make the part easier to get in spec. If you dimensioned it any other way you’d have to have extremely tight tolerances on the 45 degree angles and the overall width to prevent stacking errors in the critical locations. Believe me I’ve dealt with my fair share of dumb dimensioned parts, but knowing what mounts to these rails and how they mount. That rectangle in the middle where everything is dimensioned is very critical. There’s nothing wrong with this drawing or where it’s dimensioning from.
Occams Sawzall - Occams Sawzall - I understand your point but rail-grabbing accessories deform to clamp the rail anyways. If I were making, or doing QA on finished parts, it would be far easier to measure the actual critical dimensions of a part. If that reference box is so absolutely critical to the design, good luck measuring it with a mic.
R.C. Whitehead I explained elsewhere how it can be measured with the print dimensions manually. Parts and prints are not engineered and dimensioned for the ease of the machinist. They’re are engineered and dimensioned for the sake of assembly. Say you made the rail and mount out of titanium or inconel or 316 stainless vs aluminum. Those materials will not deform the same or possibly not enough to mate correctly if you alter the way it’s dimensioned.
I thought about that on some that I modeled up, but chose to make it max material condition and hit the tolerance either in CAM or with cutter comp out at the machine. I'd rather the first one come out oversize in other words. a GO-NOGO gage would be the way to go for production.
A subtlety of that drawing is that the four corners of the rectangle are not positioned on the mid points of the diagonal lines. Not a lot of people get that, but your model is not right because of that. It's really stupid the way it is defined, to have a box drive the dims and not have the corners on the midpoints the way that it looks at first glance. That was an idiot government engineer who wanted to show off how fancy his GD&T knowledge was.
You’ve carefully modeled the nominal dimensions while completely ignoring the specified tolerances. Since these happen to be mostly one-sided tolerances, you are on dangerous ground. You’ll need to pay attention to the tolerances when you design the toolpaths, and again when you measure the resulting part for acceptance. Depending on the rest of the workflow, you might be better off taking half (or some other fraction) of the tolerances off of your nominal dimensions.
.164 is nominal. It’s a unilateral dimension not a bilateral dimension. Every dimension on a print that’s written is treated as nominal no matter which way the tolerances go. You don’t just split the tolerance on the print to make it bilateral because you feel like it. Its dimensioned that way for a reason, and the part should be modeled to the nominal dimensions of the print.
gredangeo No. You DO NOT. Learn GD&T and what that M in a circle means. You model the part to nominal *always*. The numbers on the drawing are the nominals. Period. That’s what the part is to modeled and machines to within the specified tolerance.
So I get from a GDT standpoint it's important to model the nominal dimensions, but as someone who does the modeling, Cam and then the machining I generally model everything to the center of the tolerance. If it's so wrong to model it that way, how do you do it the "right" way and make good parts?
I would encourage using a symmetry axis and dimensioning over it.
Definitly using the a standing dimension and dividing it is an Okay way but using an equal constraint and one dimension will look cleaner in detailed scetches.
I like your information density in your tutorial. Keep up the good work.
my most favorite method for this is to project in the appropriate origin axis then toggle it to construction. fully constrained in about 3 clicks!
I was just thinking today about making a rail so I could pre-mount scopes. Totally awesome, thanks John.
Great video, from the subject, information, and process perspectives. Almost takes me back to a time when you did a lot of firearm related work on your channel. Thanks again!
Thank you very much! i was looking for a way to draw and so confused just based on the blueprint/dimensions google provided... THIS IS AN AWESOME WAY TO EXPLAIN IT KEEP IT UP!
I guess there is another video shows the rest of process : groove and drilled hole?
Great video, can you please do a second video on how to add the grooves and drilled holes
John, I have been working on parts for my pcc, I have been struggling with getting the model for the rail correct. Your video was extremely helpful! I see now what I have been missing. Thank you!!!!
There's a new standard now called the NATO Accessory Rail STANAG 4694, very similar to the MIL-STD-1913 with a couple tweaks for how accessories interface, and a big difference in that it's defined in metric. Food for thought!
Thanks for this, I used STANAG 4694 in combination with this video to fix my model.
This helps a lot. I had the idea to machine one of these for myself in school
At 3:18 you could say that the center point of the line is vertical with the middle of the rectangle to avoid another dimension
Wow. I really am having trouble with drawing in 360 after mastering solidworks over the last 20 years. It's a real challenge making the transition between platforms. I can create this in under a minute and would have gone with a center/mirror line feature as a driving start. Nice work. Thanks for sharing your knowledge.
I've been working with Solidworks for some time now
It's fun to see how different softwares use pretty much the same methods of construction.
arg, you make it look so easy 😁
you, my brother from across state borders, you have just earned yourself a new subscriber. email me one day when you get a chance. would love to learn more about cnc.
Each video I watch I get impressed by Fusion 360.
Any videos about lathe work in Fusion360 in the near future? Greetings.
I don't use Fusion360 (one of the last holdouts!) so I've been skipping the Fusion Friday videos lately but this one caught my eye. You sure made it look easy. There should be an online library of crowd sourced common parts like the Picatinny rail.
Here's one: www.grabcad.com
GrabCad
Yes +1 for grabcad
I did all the steps, and I have more space on the bottom of that hexagon shape i created than I do on the top. I followed the steps carefully, and it didn't work for me.
Hi do you have the second part on the top?
I would really like to know your method for quickly placing the slots in the top of the rail John?
Draw one slot, and then use a linear array. 🙂
i ran 28 of those last week!
I wish you could do this for the Holosun 509T optic cut for pistols.
I can't seem to find the info anywhere.
Nice example, thanks! Now I need to figure out what the heck a Picatinny Rail is :)
en.wikipedia.org/wiki/Picatinny_rail here you go, it's the rail on guns for scopes and tactical flashlights.
Ahhhh Guns... Thanks, I was just about to ask Google what she thought. I get a pass on not knowing that one - I'm Canadian, we prefer Beer, buying 50 year old clunker submarines from other countries and relying on our big brother "downstairs" to take on our fights - and we appreciate it... ;)
so darn cute, eh.
Where is the rest of it? the groove and attachment holes?
At 5:00, you can instead use the "equal" constraint on the two small edge break lines. That original drawing looks a lot more complicated than it needs to be. Is there any reasoning from a CNC perspective to make the dimensions like that?
Well its an old drawing they tend to be more complicated than what they should be. but it contains all the tolerances needed as well. i work in a really old factory its funny to read 50 year old drawing compared to 1 year old drawings.
I would have dimensioned it as a rectangle with four chamfers but maybe tolerances would stack too much. I guess it matters which faces the accessory indexes on.
That drawing isn’t complicated.
It’s dimensioned in such a way to utilized GD&T and to ensure that no matter where within the tolerances given that the mating parts will fit correctly.
How do you measure the dimensions after the part is made? For example how do you measure the 0.748 length of the rectangle with calipers? Do you need a robotic probe just to see if it meets spec?
James Patrick
It’s true position call outs would pretty much require a CMM to do it.
You can do it the old school way with V-blocks, height gauges, gauge pins and such as well. But a CMM would be the best way.
If you wanted to measure that .748 you would set 2 pins of equal size (I think .25” in this case) to a center to center distance of .835, set the part upside down to the drawing orientation, then do the math (which I’m not gonna do here) to figure the height distance from the pin center line to the flat (with the .005 flatness call out). That will get you a derived dimension that if met will confirm the .748 dimension is correct.
Excellent demo! Let’s throw a kink in the cog. How would I model 20 MOA cant into the rail over a given distance? For example, my LaRue OBR has a 20 MOA cant starting at the back of the rail and continuing into the hand guard. Do you manually do the trig and input it as a slope or can you do it in the dimension as a calculated value?
Just model it as John did, but to the length you want. Then create a second sketch on the side of the angle you want and extrude it. 20 MOA is just 1/3 of a degree
You really need to do a second part on how to finish it
Do you have any examples of how you start/draw a multi component model? For example I am currently trying to sketch up the Gingery shaper based off of prints in the book in fusion. How do you manage all components, when to do the assemblies/joints, how about fasteners? New components for each one. Sorry total newb here.
Nate Van Bynen new components for each part. You can then export all components as individual components. Then start a new file and bring in each individual part you need to do your assembly and do your joints and such there.
Fusion handles assemblies, especially large ones rather poorly. It can be done and have seen injection mold assemblies done in fusion, but it’s rather clunky. The software is clearly built more around the part side of things rather than the assembly side of things.
Occams Sawzall thanks so much for the reply, that makes more sense. I have been tryin to do joints as I go along and after a bit and a few components later I find it just gets to be to much
Its Friday somewhere!! lol
Im drawing one right now, lol!
I made one before as well or at least the model
Can you post a link to the PDF you referenced?
Done! Added to description and at www.nyccnc.com/picatinny-rail-drawing-3d-model-fusion-360-ff113/
Also google M1913 STD
I meant to say go to nyccnc.com :)
Nice tewt mate, thanks for sharing but I won't take the L plates off just yet ;-)
I guess I am Mr Nit picker this time, but the Picatinny rail was not done - you didn't do the slots!!! Great video as usual, however. Not having Fu360 I just did it straight in Gcode - a suggestion perhaps? :~>
None of this works for me…
IMHO... something is lost where a standard drawing has dimensioning made from arbitrary reference geometries rather than geometries of the actual part itself. There is a disconnect from the reality of how it would be machined on a manual machine. In my line of work (railroad track) see updated spec drawings for rail profiles and there are dimensional errors on them where transitions between fillets are not tangent to each other at osculating points because some kid fresh out of engineering school, with no background in mechanical drafting, made a drawing however it was easiest to him in CAD.
R.C. Whitehead Been fighting things like that all my life....Engineers driving up the costs of things because they don't realize the ramifications of their incompetence. Try as I might I've tried to train them in some common sense techniques but Every time they take a lunch break it's time to start over. :(
Cad-Cam_Man - I had a mechanical drafting class my freshman year of engineering school. I was perplexed how something simple like orthographic projections and isometrics made absolutely no sense to the females in the class. They had no sense of how something is constructed from 3 reference views. I bet they are now working at large firms with 6 figure salaries.
R.C. Whitehead
It’s because that’s the critical locations for the mating parts.
And the true position tolerances at maximum material actually make the part easier to get in spec.
If you dimensioned it any other way you’d have to have extremely tight tolerances on the 45 degree angles and the overall width to prevent stacking errors in the critical locations.
Believe me I’ve dealt with my fair share of dumb dimensioned parts, but knowing what mounts to these rails and how they mount. That rectangle in the middle where everything is dimensioned is very critical.
There’s nothing wrong with this drawing or where it’s dimensioning from.
Occams Sawzall - Occams Sawzall - I understand your point but rail-grabbing accessories deform to clamp the rail anyways. If I were making, or doing QA on finished parts, it would be far easier to measure the actual critical dimensions of a part. If that reference box is so absolutely critical to the design, good luck measuring it with a mic.
R.C. Whitehead
I explained elsewhere how it can be measured with the print dimensions manually.
Parts and prints are not engineered and dimensioned for the ease of the machinist. They’re are engineered and dimensioned for the sake of assembly.
Say you made the rail and mount out of titanium or inconel or 316 stainless vs aluminum. Those materials will not deform the same or possibly not enough to mate correctly if you alter the way it’s dimensioned.
I would have done half and simply mirror it. then push it out. 5 min.
I would have drawn everything up at mid-limit.
I thought about that on some that I modeled up, but chose to make it max material condition and hit the tolerance either in CAM or with cutter comp out at the machine. I'd rather the first one come out oversize in other words. a GO-NOGO gage would be the way to go for production.
so small.
Fusion Thursday? LOL
tzampini it's Friday here.... 😉
A subtlety of that drawing is that the four corners of the rectangle are not positioned on the mid points of the diagonal lines. Not a lot of people get that, but your model is not right because of that. It's really stupid the way it is defined, to have a box drive the dims and not have the corners on the midpoints the way that it looks at first glance. That was an idiot government engineer who wanted to show off how fancy his GD&T knowledge was.
3rd!
Keep up the great work John!
You’ve carefully modeled the nominal dimensions while completely ignoring the specified tolerances. Since these happen to be mostly one-sided tolerances, you are on dangerous ground. You’ll need to pay attention to the tolerances when you design the toolpaths, and again when you measure the resulting part for acceptance. Depending on the rest of the workflow, you might be better off taking half (or some other fraction) of the tolerances off of your nominal dimensions.
Using the 'coincident' function feels like cheating. 😏
1st!
It is not .164 it is .164 -.020 and same with .835 -.005
Nominal dimension is 0.164", -0.020" is the tolerance (undersize only).
no man the -0.020 is tolerance
.164 is nominal. It’s a unilateral dimension not a bilateral dimension.
Every dimension on a print that’s written is treated as nominal no matter which way the tolerances go.
You don’t just split the tolerance on the print to make it bilateral because you feel like it. Its dimensioned that way for a reason, and the part should be modeled to the nominal dimensions of the print.
gredangeo
No. You DO NOT.
Learn GD&T and what that M in a circle means. You model the part to nominal *always*.
The numbers on the drawing are the nominals. Period. That’s what the part is to modeled and machines to within the specified tolerance.
So I get from a GDT standpoint it's important to model the nominal dimensions, but as someone who does the modeling, Cam and then the machining I generally model everything to the center of the tolerance. If it's so wrong to model it that way, how do you do it the "right" way and make good parts?