3D cad design sequence

[johnmcc69]

Dude, you have Pro-E and you draw gears?! Use these (they're Wildfire 4). Make a copy or back-up for a new gear (I save these masters in my "Bolt_Bucket" folder under "Gears"). Open and go into parameters. Enter the DP (Diametral Pitch), N Number of teeth, PRESSANGLE (pressure angle) & CLEARANCE then "regenerate". Voila, your raw gear. After that you can play with the width, cut spoke holes etc.

Yep. I do.

I think it's a fundamental exercise in design. I remember drawing gears on paper when I was in tech school. I learned about the addendum, dedendum, single depth...ETC. you have to know these things if you're a designer/engineer. I can also download models off the internet, McMaster-Carr has tons of them. Gears, hardware...I've used their files for hardware (STP). Pretty. Well detailed right down to the helix angle. I gave that up as well, my reason being this: (aside from the fact that my printer makes those fancy threads as blobs & wasted ink..) I can have ONE "Family table" part that has ALL the screw sizes I need, along with the proper parameters that will show up in my BOM & am able to replace those at the click of a mouse without having to download a file from the internet. I.E.: I have a file called 1-4-SHCS,, it has all the length screws I would ever use. Of course, all of these have proper parameters & will update my drawings BOM as I mentioned.

I apologize if I derailed too far here. I don't work in a "production" environment anymore & time is not a factor, it's just a hobby now.

I do really appreciate the files & will have a look at them, it will be fun to play with.

John

 

[L98fiero]

LOL, yes, the Kaypro. I knew someone who had one and I drewled all over myself wanting one of those. I evcentually got my first computer in 1977--a Radio Shack TRS-80. Slowly bought th3e parts that made a real computer with a zinging 4MB clock, two floppy drives (each 500$) and a printer (1700$) and what a clunker it was! In the end, it cost a total of 5000$ in 1977 dollars. I estimate that to be somewhere around 35000+ $ in our money today. I mean you could buy a loaf of bread for 25c back then. I would be reading magazines that advertised a COLOR computer for 13000$ which I also drewled over. It had 4 colorsw. In '88, I got my second computer which was an Amiga--for the $ it was the berst computer ever made. Had they stayed in business, I thimpfks it would be the best today if they had kept up with all the innovations. I still have that Amiga. There are Amiga clubs today, did you know that?

Richard, about the time you got your Amiga, I got an Atari 520ST, they had issues with Apple because they more or less copied the MAC graphical UI, with Lotus 123 and, I can't recall the name of the CAD program, but there was no typing point by point to get a line like ACAD, it was graphical UI and a mouse, a lot like my 1990s TurboCad 16.2 that I still use for simple 2D drafting. I never did like ACAD. Oh, and it had a, pretty good for it's time, 2½ axis CAM program, I think it was called NC-LINK.

 

[oldCB]

The experts say that the body is better staying on standard time. For me I would just like to leave it alone. I can live with either one, just quit changing it.

Could not agree more. Standard Time For All !!!

 

[Richard Hed]

Could not agree more. Standard Time For All !!!

You've got to remember, Daylight Savings Time was a WAR measure to save precious energy. We are still at war everywhere on the globe so I guess we still needs the DST

 

[stanstocker]

Not sure how we got onto DST/STD time here, but standard time? There are 24+ of them around the globe... Not to mention sunrise and sunset move 4 minutes for each 100 miles. Only reason I like DST is that I'm more of an evening / night guy, so like having the light "later". Silly I guess as time as we measure it in the conventional (12 hours twice a day, 60 minutes and hour, not Planck time) sense is purely a man made artificial construct anyway.

 

[Gordon]

I am sure that a majority of folks would be happy just to leave it one way or the other. Not to get too political but why does this not get done? The legislature has to change it but they are too deadlocked to do ANYTHING. There are a lot of similar issues where the public is overwhelming in favor of something and nothing changes. So much for the legislature is representing the people.

 

[Richard Hed]

I am sure that a majority of folks would be happy just to leave it one way or the other. Not to get too political but why does this not get done? The legislature has to change it but they are too deadlocked to do ANYTHING. There are a lot of similar issues where the public is overwhelming in favor of something and nothing changes. So much for the legislature is representing the people.

Where I work, the company just makes up their own time to show up.

 

[Gordon]

Where I work, the company just makes up their own time to show up.

Me too but I am retired. The big problem is management (AKA wife) does not agree with my desired times. Another political problem.

 

[Richard Hed]

Me too but I am retired. The big problem is management (AKA wife) does not agree with my desired times. Another political problem.

LOL

 

[awake]

You won't be able to change individual items like the bore or boss as if you were editing a gear you had modelled yourself as they are just imported as a solid.

So think of it as a purchased gear and sketch the things you would machine.

Want a bigger bore then sketch a circle on one face and cut-extrude through all
Want a smaller bore then draw 2 circles and extrude those to "sleeve" the existing bore
Bigger boss draw a bigger circle and extrude
Smaller boss circle what you want to remove and cut-extrude just like turning down the boss
Shorter boss just draw a rectangle with one edge dimensioned from the edge of the existing boss by how much you want it shorter and cut-extrude

I had no problem doing any of those things with one of the gears from your link imported as a STEP file. If you look at the image I posted yesterday where I made the gear a lot wider you can see the red sketch which was used to sleeve down the bore and there is a boss on the opposite side to the extrusion of the gear profile.

When I imported the file it was concentric to the Z axis and as the gear is a single part should not need moving further. If you want to put it into an assembly then just the usual concentric constraint of the gears bore and the face of a shaft will do.

That's pretty much the way I do it. I'll model a single tooth (It doesn't have to be fancy, but the geometry is pretty simple as are the tooth calculations.) & as Richard says, make a circular pattern of that tooth profile.

When you try to import all those segments from "Imported geometry", you may end up with an "unconnected" sketch (I don't know what it's called in different softwares), but, I know in my experience, I've spent a lot of time trying to close up & fix a sketch like that.

In Pro-E, I can "toggle" a sketch to show the interior sketch as shaded, or "filled", if it is indeed a closed sketch & usually don't have any problems. Does Alibre have something like that? Gordon also mentioned problems with the extrude direction not being able to predict the direction: Does Alibre have any kind of "Preview" feature that shows you the extrude before exiting the command?

I've usually had pretty good luck with the Rush Gear STP files, but I usually modify them exactly as Jason mentioned, but...I find it's just as easy to create the single tooth & pattern that. You can then take that a step farther & save a sketch (As its own "sketch" file, properly dimensioned & (ahem) Constrained, & use that tooth profile in any gear design). A caveat with that method may be regeneration times & file sizes, but depending (again) on your software, you may be able to "group" or "merge" all the teeth together or even save out your gear file as a STP file, import it back in, & then save that as your model file, it will then come in as one imported solid feature & not have to regenerate all the tooth profiles. If you're not happy with that gear, you can go back to you original gear part file (You didn't delete that right?) & make your modifications & then export it again as a STP. this method seems to work pretty well with model files that have a lot of patterns & complexity.

I hope this makes sense & helps.

You're doing just fine Gordon. Keep at it.

John

Dude, you have Pro-E and you draw gears?! Use these (they're Wildfire 4). Make a copy or back-up for a new gear (I save these masters in my "Bolt_Bucket" folder under "Gears"). Open and go into parameters. Enter the DP (Diametral Pitch), N Number of teeth, PRESSANGLE (pressure angle) & CLEARANCE then "regenerate". Voila, your raw gear. After that you can play with the width, cut spoke holes etc.

Yep. I do.

I think it's a fundamental exercise in design. I remember drawing gears on paper when I was in tech school. I learned about the addendum, dedendum, single depth...ETC. you have to know these things if you're a designer/engineer. I can also download models off the internet, McMaster-Carr has tons of them. Gears, hardware...I've used their files for hardware (STP). Pretty. Well detailed right down to the helix angle. I gave that up as well, my reason being this: (aside from the fact that my printer makes those fancy threads as blobs & wasted ink..) I can have ONE "Family table" part that has ALL the screw sizes I need, along with the proper parameters that will show up in my BOM & am able to replace those at the click of a mouse without having to download a file from the internet. I.E.: I have a file called 1-4-SHCS,, it has all the length screws I would ever use. Of course, all of these have proper parameters & will update my drawings BOM as I mentioned.

I apologize if I derailed too far here. I don't work in a "production" environment anymore & time is not a factor, it's just a hobby now.

I do really appreciate the files & will have a look at them, it will be fun to play with.

John

I'm intrigued - do neither Alibre nor Pro-E have gear profiles built in? In FreeCAD, you can input the DPI / module and number of teeth, and instantly get a sketch of the gear, which you can then extrude. (You can also select that this is for an internal gear, e.g. a ring gear, and it will adjust the shape accordingly so that you can cut or pocket instead of extruding.)

I've only ever used this with spur gears ... can't remember if it can also generate other types of gear. That may require an add-on, but if so, the add-on modules are free!

 

[Jasonb]

I did mention it in one of my posts but Alibre Pro and Expert have a script for generating gears and is what I tend to use, not sure if it is in Atom but I don't think it is. F360 also has "FMGear" for generating gears which also does helical ones

 

[awake]

I did mention it in one of my posts but Alibre Pro and Expert have a script for generating gears and is what I tend to use, not sure if it is in Atom but I don't think it is. F360 also has "FMGear" for generating gears which also does helical ones

Ah, sorry - I missed seeing that.

 

[Jasonb]

I've been swapping a few PMs and e-mails with Gordon, mostly to do with Assembly within Alibre and did a video of the early stages of assembling the vee twin that he is working on using the parts he had modelled. I'm sure he won't mind me posting it here if it helps anyone else playing with Alibre. I'm using pro so one or two things may look a bit different but I'm 99.9% sure what I show is available in Atom too. You tube playback speed can be altered if you can't keep up.

 

[Zeb]

Something I don't think was mentioned but thoroughly elaborated on is design intent. It is maybe an overused industry term shrouded in black magic.
You basically sit down with a project and define how parts will be related to each other with planes and sketches. Then you look at how those might change over time, so editing in the future is easier. Everyone is different and opinionated, so it is what works for you.

You can get fancy with formulas that automatically model and tolerance your entire assembly with a simple length change to the rod or crank journal.
This is not necessary. A big problem is you get a different engineer on the same project (or I'm ten years older) and all of the sudden the assembly becomes a mess with manual workarounds. The first design is never the simplest, the best designs are elegantly approachable.

Assembly datums and planes are the most important. As tempting as it might be, you might not want to start at the top of the cylinder.
For my engine projects the datum is typically the crankshaft flange, as anything can grow or shrink from that point. An aircraft might be a firewall. I know of an aircraft manufacturer that used a point in space a foot forward of the spinner (not recommended).

Also good practice is to leave form and fit for last. I'm seeing a lot of smaller steam engine projects that are underpowered for the task, so I'm looking at gearing a project way down and adjusting bore and stroke to match the output. If my datum is the crank, everything can move around it.

For gears, I only use the major Ø for visualizing collisions and pitch Ø for design. This can be just a couple of circles on a sketch. Modeling gear teeth for instagram design sauce can be left for the end. The key is to enhance clarity by removing everything in the equation that doesn't matter. Sometimes I'll just model sketches and go straight to machining.

That was quite the ramble...blaming caffeine lol.

 

[johnmcc69]

Nice work Jason & Gordon!

Sound advice Zeb.

John

 

[Richard Hed]

Hey Gordon,
We need to hear from you regularly. How are you doing on the Alibre?

Has anyone built the DOUBLE SIDED BEAM ENGINE by Bob Middleton featured in MEWS? I'm drawing it up but keep misplacing some of my drawings and have to draw it up a 2nd time. What a pain.

 

[Gordon]

Hey Gordon,
We need to hear from you regularly. How are you doing on the Alibre?

Has anyone built the DOUBLE SIDED BEAM ENGINE by Bob Middleton featured in MEWS? I'm drawing it up but keep misplacing some of my drawings and have to draw it up a 2nd time. What a pain.

I have not progressed much beyond what Jason has shown. He used the parts that I had drawn and basically cleaned up my assembly by doing it correctly. I am presently actually making the parts to give my mind a rest from the desk work. I am producing the 2D drawings from my 3D models as I need them. So far I have made 3D models of most of the major parts and have made an assembly from those parts and made 2D drawings of about 75% of the parts. I have progressed far enough to know that the program is capable of doing what I want to do. I want to actually use the program to make an engine of my own design instead of using someone else's design and duplicating it. I have determined that Free Cad is capable of more but is harder to learn. Also I like the idea of owning the program and having it living on my computer instead of doing work in the cloud. I learned some time back that something like Fusion 360 can start out free for hobby users and then suddenly require a yearly subscription and unless I pay up my past work is gone. Alibre has more advanced programs but it is hard to justify the expense for a hobby and most of the enhancements are things that I don't really need or just whiz bang show.

 

[Zeb]

Hey Gordon,
We need to hear from you regularly. How are you doing on the Alibre?

Has anyone built the DOUBLE SIDED BEAM ENGINE by Bob Middleton featured in MEWS? I'm drawing it up but keep misplacing some of my drawings and have to draw it up a 2nd time. What a pain.

I took a look at the drawings just now. What are you having difficulty with? If the project feels too overwhelming, this is how I might organize it:

I wrote it in a super overwhelming way... I'm mildly sick and not allowed to work today...so...

_____Summary

• Create the granddaddy reference plane from baseplate top surface
• Define an axis in space for flywheel zone
• Define an axis in space for valve timing finger jamming elements
• Define an axis in space for lower beam rocker pivot axis (not covered)

_______Modeling 1

• Model the base plate (A) using a center rectangle (not corner) in the sketch, that way you can reuse the existing planes for mirroring. Extrude down, not up so when you drop it at next higher assembly the top face will be zero. Might not be necessary but I like to organize that way. The top face of the baseplate would be the granddaddy of all datum feature references. Any dimension screwed up here will work its way through the assembly.

_______Modeling 2

• Model the center beam (B) and giant yoke thing (C). You can model half the yoke and mirror to save time on future updates (if symmetric).
• Model the 1x lower hanger bearing housing (D) that the flywheel assembly mounts to. The virtual center will set the axis from which everything else on the flywheel shaft will be located from.

________Modeling 3

• Block out 1x cylinder with base (E) and model the 6 holes that the vertical beams attach to. Do not model the guts or anything else.
• Model 1x vertical beam. This will set the datum axis for your valve timing thingamaggigers that will rock back and forth. Ignore those parts for now.

________Assembly 1

• Create an assembly file and set the base plate (A) with a fixed constraint.
• Add center beam (B) and constrain to (A) (concentric+touch).
• Add ref plane to where the top face of the giant yoke thing would sit. Constrain by distance to base plate (A).
• Add giant yoke thing (C). Constrain top face to plane just created.
• Add 2X hanger bearing (D) and use mounting holes and mount to yoke. You can mirror these parts later using the mirror plane in the first step.
• Create an axis datum on inner surface of any hanger bearing (D). From now on you can drop the flywheel and any other parts on this axis datum. You can use the plane bisecting the entire assembly to center parts or mirror.

_______Assembly 2

• Add 2X cylinders with bases.
• Add 4X vertical beams to cylinders. You can install 2x and mirror on the center plane if you want.
• Pick a vertical beam and add an axis datum going through all the vertical beams. You will now be able to drop all the valve timing thingamagiggers at a latter time.

______________
From there you can divide and conquer or send individual parts for people to noodle on. Subassemblies can grow or shrink from the fixtures. I didn't cover the lower rocker beam, but the same principles apply.

Bonus...
If you have a bridgeport...and are feeling adventurous and overly trusting of the engine designer, you can slightly oversize all the mounting holes where bolts will pass thru. Then add 2x dowel pins and reamed holes to accurately locate them. The screws can then snuglify and the pins alignify.

 

[johnmcc69]

What the heck are you talking about!?

No, seriously I totally understand your proposed methodology of "Top down" design & as PTC called it "Skeletons" (Using an assembly file with (only) all the important datum planes & defining axis-zzz (like that?). But, that becomes a more advanced technique & requires a little more attention to how all the assembled parts will fit. It would be easy to assemble a part to another part that isn't defined by the "Granddaddy refrence plane", & run into assembly failures-this has bit me in the arse downstream when I have tried to run animations. I don't mean just "Dragging" the flywheel around & making it go 'round, but applying a "motor" (& velocity/speed, ETC.), & running it through all its motions. (of course, this is only my experience with my CAD software (Pro-E WF4), I'm sure , it's better now..).

But as you said "...are feeling adventurous and overly trusting of the engine designer", you can use this method with confidence. I'm not that confident, I design from the "Bottom up" with the exceptions of creating parts within the assembly to maintain those references. +1 on using dowel or roll pins, it beats the hell out of trying to align many (critical) pieces at one time.

This IS a very useful technique, but like everything, it has its place. For model engines...I dunno..

Gordon! We need pix of your progress!!!

John

P.S.: Boy, that's a pretty engine, the double sided beam, if my CAD computer wasn't down, I would love to re-draw that.

 

[Richard Hed]

I took a look at the drawings just now. What are you having difficulty with? If the project feels too overwhelming, this is how I might organize it:

I wrote it in a super overwhelming way... I'm mildly sick and not allowed to work today...so...

_____Summary

• Create the granddaddy reference plane from baseplate top surface
• Define an axis in space for flywheel zone
• Define an axis in space for valve timing finger jamming elements
• Define an axis in space for lower beam rocker pivot axis (not covered)

_______Modeling 1

• Model the base plate (A) using a center rectangle (not corner) in the sketch, that way you can reuse the existing planes for mirroring. Extrude down, not up so when you drop it at next higher assembly the top face will be zero. Might not be necessary but I like to organize that way. The top face of the baseplate would be the granddaddy of all datum feature references. Any dimension screwed up here will work its way through the assembly.

_______Modeling 2

• Model the center beam (B) and giant yoke thing (C). You can model half the yoke and mirror to save time on future updates (if symmetric).
• Model the 1x lower hanger bearing housing (D) that the flywheel assembly mounts to. The virtual center will set the axis from which everything else on the flywheel shaft will be located from.

________Modeling 3

• Block out 1x cylinder with base (E) and model the 6 holes that the vertical beams attach to. Do not model the guts or anything else.
• Model 1x vertical beam. This will set the datum axis for your valve timing thingamaggigers that will rock back and forth. Ignore those parts for now.

________Assembly 1

• Create an assembly file and set the base plate (A) with a fixed constraint.
• Add center beam (B) and constrain to (A) (concentric+touch).
• Add ref plane to where the top face of the giant yoke thing would sit. Constrain by distance to base plate (A).
• Add giant yoke thing (C). Constrain top face to plane just created.
• Add 2X hanger bearing (D) and use mounting holes and mount to yoke. You can mirror these parts later using the mirror plane in the first step.
• Create an axis datum on inner surface of any hanger bearing (D). From now on you can drop the flywheel and any other parts on this axis datum. You can use the plane bisecting the entire assembly to center parts or mirror.

_______Assembly 2

• Add 2X cylinders with bases.
• Add 4X vertical beams to cylinders. You can install 2x and mirror on the center plane if you want.
• Pick a vertical beam and add an axis datum going through all the vertical beams. You will now be able to drop all the valve timing thingamagiggers at a latter time.

______________
From there you can divide and conquer or send individual parts for people to noodle on. Subassemblies can grow or shrink from the fixtures. I didn't cover the lower rocker beam, but the same principles apply.

Bonus...
If you have a bridgeport...and are feeling adventurous and overly trusting of the engine designer, you can slightly oversize all the mounting holes where bolts will pass thru. Then add 2x dowel pins and reamed holes to accurately locate them. The screws can then snuglify and the pins alignify.

You do it differently, but I like the way you have explained. I will try your method. This is not that difficult of a set of drawings, it's just that I must have accidentally over wrote some (impossible) or lost them in the maze of the files. I usually have a very good system but between backing them up I lost some. don't know what happened to them but I do know I did them then had to re-do them . That was OK as the second time thru I managed to create them far more efficiently.