3D Cad (Lesson 101)

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

mike_t_12

Well-Known Member
Joined
Mar 7, 2010
Messages
50
Location
Los Angeles, Ca
Hi Jan, I made this Bugatti racer model a while ago in Rhino 3d... it needs a lot of detail work including the prop! For reflections, the model is inside a hanger-like building with white skinney skylights. What software are you using?

Wagy59, I didn't know you could do 3d models in Autocad - it's looking good. Is the Rhino version of your twin coming soon?!

Inverted Vantage, your Mustang is looking good. I've never used Modo before but would like to try someday. The Mustang is my all time favorite airplane!

Mike
Bugatti_racer_2.jpgBugatti_racer_1.jpg
 

Inverted Vantage

Formerly Unknown Target
Joined
Jun 19, 2008
Messages
1,116
Thanks mike_t_12; that is some great surface work there, how'd you do that? Especially the fillets, what was your thought process and procedure?
 

mike_t_12

Well-Known Member
Joined
Mar 7, 2010
Messages
50
Location
Los Angeles, Ca
Thanks Inverted Advantage. I really enjoy making these 3d models. I started by tracing the top and side view from a blueprint drawing. Then, I created some fuselage curves (highlighted in yellow) to use for lofts and rail sweeps. I played around with the shape of the fuselage curves based on looking at pictures of the real airplane. So, the model is more of a concept and not a very accurate copy. For the wing fillet (highlighted in yellow), I used the "Blend Surface" command with position (G0) continuity on the fuselage and curvature (G2) continuity on the wing.
Mike
blueprint setup.jpgside view blueprint.jpgtop view blueprint.jpgfusalage curves.jpgfillet_blend.jpg
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
3,946
Location
Warren, VT USA
Thanks Inverted Advantage. I really enjoy making these 3d models. I started by tracing the top and side view from a blueprint drawing. Then, I created some fuselage curves (highlighted in yellow) to use for lofts and rail sweeps. I played around with the shape of the fuselage curves based on looking at pictures of the real airplane. So, the model is more of a concept and not a very accurate copy. For the wing fillet (highlighted in yellow), I used the "Blend Surface" command with position (G0) continuity on the fuselage and curvature (G2) continuity on the wing.
Mike
Really nice stuff. Definitely from the era of the spline curve from strips of spruce and lead weights era :) Probably why it looks so cool. I would guess that if the original conspirators could see someone doing this sort of lofting this quickly with this sort of tools it would literally blow their minds.
 

Woodenwings

Well-Known Member
Joined
Aug 2, 2006
Messages
164
Location
Toronto
Great looking models.

Here is my definition (based on Work/Education experience).
  • CADD = Computer Aided Design & Drafting. The Drafting part is not required if you mfg. parts directly from models but it is required in the certification of an aeroplane and often for many other purposes, not least the manufacturing of the product.
  • CAD is a simplification of CADD (i rarely use the CAD over the more correct CADD).
  • Parametric design techniques are just that - able to be modified later without rebuilding geometry. 2D design is so horribly slow because of the parametric "shape-shifting" is missing.

I see lots of beautiful models on this website but have never noticed any discussions of the correctness of countours and profiles. I have spent Hours fussing over 0.001 of an inch on various machined aluminum surfaces (complex curves and shapes). While i delight in the creativity and romance of creating beautiful 3D models (regardless of platform) - How important is correct geometry to most builders?

Tell us what you think...

What technical understanding/ability bridges the gaps between pretty model and flying model? I'm not looking to discourage anyone from making a pretty model. I would just like to steer the conversation towards making things correctly and in a useful way for Mfg.

If i'm building in wood i try to be within 1/32" for contours. Sometimes it is near that - i simply shrug both shoulders if the wood moves after Mfg.

How do you guys handle the nitty-gritty.:lick:
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
3,946
Location
Warren, VT USA
Great looking models.


I see lots of beautiful models on this website but have never noticed any discussions of the correctness of countours and profiles. I have spent Hours fussing over 0.001 of an inch on various machined aluminum surfaces (complex curves and shapes). While i delight in the creativity and romance of creating beautiful 3D models (regardless of platform) - How important is correct geometry to most builders?

Tell us what you think...

How do you guys handle the nitty-gritty.:lick:
First, CAD: computer aided design; CADD, computer aided design and drafting; CAM: computer aided manufacturing... Hardly anyone uses the term CADD anymore. And on top of that the attitude that CAD somehow makes things more accurate is all over marketing but not true generally. CAD models are cartoons unless there is a team to get it into reality properly.

Segue into the other question. Accuracy is about manufacturing techniques. Those can be splitting a pencil line with a table saw or cutting a 5 axis profile with a CNC machine. So the answer like all else is "it depends."
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
It's the nitty-gritty that can make you pull your hair. And it's really a challenge sometime when working in metal. Wood and composites have a bit more flexibility but in metal, even temperature changes can affect your assembly. The trick then is knowing where you can be "flexible" and where the work has little to no give.

The biggest problem the industry has is computer designers/operators who have only book knowledge but no hands on experience. The mainstream aerospace industries are always fighting that battle: Engineering departments developing beautiful models that unfortunately cannot be built.

In our work (where I do most of the design and fabrication) it's not that difficult to do since I know that I myself will pretty much have to be the one either building or helping build, especially if I design something into the proverbial corner. Fortunately it doesn't happen too often. The key then is trying to work with the design as if you were physically building it in the computer. As long as you maintain that thought process, the problems can be minimized.

We really have little to no use for parametrics. I've tried it a few times in the past (had access to Pro/E) but in this type of work I usually know exactly how things have to be arranged or designed so if I do have to do any changes, they are well beyond what parametrics could do for me. As such, they would most likely only extend the time to finish a particular set of tasks rather than aid the project.
 

Inverted Vantage

Formerly Unknown Target
Joined
Jun 19, 2008
Messages
1,116
First, CAD: computer aided design; CADD, computer aided design and drafting; CAM: computer aided manufacturing... Hardly anyone uses the term CADD anymore. And on top of that the attitude that CAD somehow makes things more accurate is all over marketing but not true generally. CAD models are cartoons unless there is a team to get it into reality properly.

Segue into the other question. Accuracy is about manufacturing techniques. Those can be splitting a pencil line with a table saw or cutting a 5 axis profile with a CNC machine. So the answer like all else is "it depends."

On this note, yea; a CAD model is just like a regular physical model. There's different levels of complexity and accuracy, depending on what you're making and what information you want to glean from it. If all you want to check is the general shape, then sure throw it together in an hour and see how it looks from different angles. If you want to check the interference of two panels, then you're going to have to significantly more in-depth. I find a lot of the time this sort of detail radiates; if you're checking the fidelity of the wing-root/fuselage connection, then you're going to have to make the internal structure and panel thickness of that area be pretty spot on. But you don't need to put in ailerons or detail the cockpit canopy latches; so models end up having detail clumped together.

That is, unless you're developing the whole thing in SolidWorks or somesuch; then yes you can have the entire aircraft down to machine-capable geometry. But you're probably going to go through three or four iterations before you reach that point, simply because as you develop it in 3D you start finding some things don't work out; this is where (and this is a segway into orion's comment) hands-on experience becomes essential. I don't have as much hands-on experience as I would like, so sometimes I'll put something into the model that can't be built; other times I'll be messing around with crazy geo and I'll realize that something simply won't work in reality; or even on the computer. On one project, I had this one curved surface in mind that blended three different ones together. It took me two days to figure out how to do it, and by the end it was pretty much a hackish workaround; so I changed the design. That's another thing; when working in CAD, you not only have to have an appreciation of the physical world, but an appreciation of the digital and the rules of the software.

Unfortunately, not enough schools have a required model airplane building course. This would really alleviate a lot of these experience issues, I feel, as people get hands on analogues to what they're designing on the computer. But orion's also correct; what helps a lot, especially in engineering applications, is to imagine the processes to build the physical model as you build the digital one; though sometimes I have found this to arbitrarily limit me and my thought process. Again, it's about understanding the material, the model, and yourself.

Parametrics are a mixed blessing; on the one hand, they're great for small adjustments or quick testing of large ones (for instance, to run through different aircraft layouts, you can create a couple parametric sketches in SolidWorks and, by using a design table, dynamically alter the design based on any set parameter), but it's not like you can design the plane in wood and then switch it to metal without running into problems. The requirements of parametric software can also slow down the design process, as they require a lot of forward planning and most of the time require you to have most of what you're building already figured out before you build it; though as you gain experience you can use it for messing around more.
 

Woodenwings

Well-Known Member
Joined
Aug 2, 2006
Messages
164
Location
Toronto
Seeing as we all agree - perhaps we should discuss the nitty gritty as it comes up.
i'm sure most people here simply enjoy making pretty models (no harm in that).
But perhaps we can talk about how to properly organise files and make parts for real.

I get nervous when people are using rhino and then say they are going to machine billet Al. from it.
 

Woodenwings

Well-Known Member
Joined
Aug 2, 2006
Messages
164
Location
Toronto
OK educate me then,

How do you specify tolerances and keep records of revisions and such? can you do prelim. mechanical analysis?

do you make drawings of parts for your records?

i'm not saying it isn't possible (obviously if you are doing it already!).

do you do that stuff?

:)
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
Yes, all procedures for making parts can be done within Rhino. All surface and solid parts are made as part of the overall airframe and/or assembly model. Components are then exported either as Rhino files or *.STP files (depending on which machine shop we're working with) - quite a few of our suppliers deal directly with Rhino. All 2-D drawings are also made in Rhino although these are primarily for annotation and tolerancing. 2-D drawings are also used to identify assemblies and respective components, as well as provide notes for the archives. Manufacturing part tolerances are specified as necessary in the 2-D drawings.

Since we primarily work with prototypes and one-off designs, revisions are rare however, in those cases we only keep the latest version. Archived drawings (if any) are numbered by date of creation and revision numbers are listed in the title block.

Structural analysis is done within Rhino (for metal parts - Scan-n-Solve). Assemblies and laminate parts are analyzed in a standalone version of COSMOS/M.
 

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,268
Location
CT, USA
When I was doing mold tooling design for automotive composite structures, we used a mix of Cadkey, Rhino, a couple of other packages (engineer's choice, since we sent the moldmaker neutral format files, IGES if I recall). There were few or no actual drawings except for some secondary operations since the shapes were too complex; the CAD surface data was the control. Tolerances were given as maximum deviation from the surface data (and in most cases were well within the moldmaker's ability). We transmitted the CAD data to the moldmaker with a cover sheet giving the file name, size, and date, that was the extent of our revision control.

-Dana

A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects.
- Robert A. Heinlein
 

Inverted Vantage

Formerly Unknown Target
Joined
Jun 19, 2008
Messages
1,116
You know what would be really useful, for this forum and for posterity? A library of 3d aircraft models, maybe three or four of them, with all the control linkages, internal structures, etc, modeled in various levels of detail. It'd be a great resource for teaching and learning aircraft design. I can't contribute right now but maybe some other members could. :)
 

Wagy59

Well-Known Member
Joined
Nov 25, 2011
Messages
270
Location
Farmers Branch, Texas
It's the nitty-gritty that can make you pull your hair. And it's really a challenge sometime when working in metal. Wood and composites have a bit more flexibility but in metal, even temperature changes can affect your assembly. The trick then is knowing where you can be "flexible" and where the work has little to no give.

The biggest problem the industry has is computer designers/operators who have only book knowledge but no hands on experience. The mainstream aerospace industries are always fighting that battle: Engineering departments developing beautiful models that unfortunately cannot be built.

In our work (where I do most of the design and fabrication) it's not that difficult to do since I know that I myself will pretty much have to be the one either building or helping build, especially if I design something into the proverbial corner. Fortunately it doesn't happen too often. The key then is trying to work with the design as if you were physically building it in the computer. As long as you maintain that thought process, the problems can be minimized.

We really have little to no use for parametrics. I've tried it a few times in the past (had access to Pro/E) but in this type of work I usually know exactly how things have to be arranged or designed so if I do have to do any changes, they are well beyond what parametrics could do for me. As such, they would most likely only extend the time to finish a particular set of tasks rather than aid the project.
Yep...you be right on..I agree
 

Wagy59

Well-Known Member
Joined
Nov 25, 2011
Messages
270
Location
Farmers Branch, Texas
First, CAD: computer aided design; CADD, computer aided design and drafting; CAM: computer aided manufacturing... Hardly anyone uses the term CADD anymore. And on top of that the attitude that CAD somehow makes things more accurate is all over marketing but not true generally. CAD models are cartoons unless there is a team to get it into reality properly.

Segue into the other question. Accuracy is about manufacturing techniques. Those can be splitting a pencil line with a table saw or cutting a 5 axis profile with a CNC machine. So the answer like all else is "it depends."
I agree and disagree with both of ya....plain ole Autodesk autocad is terrific for a starting point. anyone that says it isn't or is not suited doesn't know know what they are talking about..no matter what you start with, whether it's a pencil and markings on a floor, or the latest high end fancy software, it all comes down to making parts and trying to get them to all fit together and create something that looks and acts like an airplane..geez...forget all the design/engineering software crap..it wont build anything at all for you..
 

Inverted Vantage

Formerly Unknown Target
Joined
Jun 19, 2008
Messages
1,116
I agree and disagree with both of ya....plain ole Autodesk autocad is terrific for a starting point. anyone that says it isn't or is not suited doesn't know know what they are talking about..no matter what you start with, whether it's a pencil and markings on a floor, or the latest high end fancy software, it all comes down to making parts and trying to get them to all fit together and create something that looks and acts like an airplane..geez...forget all the design/engineering software crap..it wont build anything at all for you..
Well no software claims to do all the designing for you. If it could, then it would be sentient, now wouldn't it? A piece of software is like a hammer or a wrench; it's just a tool.

AutoCAD is pretty outdated and is very much a legacy program IMO. It has a wicked outdated user interface and it's 3d capabilities, while I've seen people do amazing things with them, are not as easy to use as many other software out there at half the price.
 

Wagy59

Well-Known Member
Joined
Nov 25, 2011
Messages
270
Location
Farmers Branch, Texas
Well no software claims to do all the designing for you. If it could, then it would be sentient, now wouldn't it? A piece of software is like a hammer or a wrench; it's just a tool.

AutoCAD is pretty outdated and is very much a legacy program IMO. It has a wicked outdated user interface and it's 3d capabilities, while I've seen people do amazing things with them, are not as easy to use as many other software out there at half the price.
91.jpg
Does that mean I'm just unusually **** good at it despite the fact I ain't dat smart??


You might inform Autodesk and the umteen tens thousands of companies that still use it everyday for engineering , structural, and yes of course architectural work and pay a hell of a lot of money in licensing... that they are all in the stone age and doing it all wrong with outdated, archaic, lousy software they just paid thousands for......LOL Don't get me wrong..I ain't an autodesk fan at all..
 
Top