T-Splines and model aviation

[aeronautica]

This is my first post in this forum.
First, a little background of what I want to be able to do with Rhino and T-Splines.

I am an architect and I have been using Microstation since 1982, when it was an Intergraph product, in my architectural practice. During this time, I have only been using the software for 2d drawing.

My hobby is large-scale, gas powered, r/c planes that I fly in competitions. The area of the hobby that I am involved with is “Scale”, models of actual aircraft. The sizes of my planes are from 1/5 scale to 40% of the full size planes. I have decided to start up a new business in designing and laser cutting kits of my own designs of various full size aircraft.

Working within Microstation I have drawn detailed 2d 3-views a Cessna 152 that is my first project. I then have used Compufoil3d software for the design of the wings and tail surfaces and I have exported this info as dxf files into Microstation. I then make my detail rib drawings that I use for my laser cutter.

The fuselage is the problem area that I hope Rhino and T-splines can help me with.

Question #1 I can import my detailed dgn files into Rhino. I was looking at using the box modeling method and use these files as reference, or can they be used for another method?

Question#2 When I import my dgn file into Rhino they are just a fraction of the size that they were in Microstation. Is there some way to scale these up or is there a setting within Rhino for this?

As I work on this project I know many more questions will come up.

I have seen the posts here and on R/C Groups of what people are able to do with Rhino in designing model planes. A lot of their work is for small foam electric planes but I am after the workflow process to produce large scale, museum quality, flying r/c models that I can design and kit to the modeling community. Accuracy of the aircraft outline is extremely important to maintain thru-out this process. This Cessna 152 will have a wingspan of 120” and weigh about 35 pounds.

Attached is a screen shot of my dgn file with the 3-vews set as a reference file that I work over for the wing a tail design.

SDS has posted a P-51 that looks great and makes me think that T-Splines is the way to go about this. My question to SDS is how did you setup your reference files to do this?

My knowledge of 3d, Rhino and T-splines is very limited and I am open to any and all ideas about how to use this software to achieve what I want to do.

[SDS]

The drawing is brought into Rhinceros using the "Picture Frame" function, scaled, leveled, then sliced into portions to arrange side, top, and front views in 3D space. Lock the drawing layer and make the drawings transparent so the model can be edited without disturbing the drawings.
Congratulations, going into the model business is a lofty goal; it seems to be a pretty crowded field.

[JSantocono]

Hi aeronautica,
Welcome to the forum.

Question #1 I can import my detailed dgn files into Rhino. I was looking at using the box modeling method and use these files as reference, or can they be used for another method?

Yes, you can import .dgn files into Rhino to use them as reference, or to use the actual geometry as starting point for the 3D surface. I think that its easier to make it using the geometry as reference and box modeling tecnics, but if you are looking for extreme accuraccy using the actual geometry as starting point would be the best, but tricky.

Question#2 When I import my dgn file into Rhino they are just a fraction of the size that they were in Microstation. Is there some way to scale these up or is there a setting within Rhino for this?

hmm....not sure. I dont have experience with .dgn files. But what I would do is to: First, set the same units on both softwares before exporting/importing. Then, do the maths to find out how much to scale (use the _Scale command).

As I work on this project I know many more questions will come up.

Feel free to ask

JuanS

[petertha]

I suspect your scaling issue is a relatively simple fix, either importing it into a specific rhino template, or, once in rhino you just need to scale it to your desired size using rhino’s button-click commands. You can do it in 1D, 2D, 3D… all kinds of tricks & sophisticated manipulations that will assist when your reference material is distorted... and it very likely is distorted if it started life in the paper world. It might be quicker to search (free) rhino support forum or post a question, lots of capable experts there http://news2.mcneel.com/scripts/dnewswe ... roup=rhino

...My hobby is large-scale, gas powered, r/c planes that I fly in competitions....
... A lot of their work is for small foam electric planes but I am after the workflow process to produce large scale, museum quality, flying r/c models that I can design and kit to the modeling community....

I am confident with a bit more searching you will find examples rhino work that will satisfy your requirements. Here are some you may find of interest:

a recent rhino/rc scale aapplication
http://www.rcgroups.com/forums/showthread.php?t=1153284

the mother of all RC composite construction projects demonstrating rhino designed surfaces exported to make cnc plugs, molds etc
http://www.rcgroups.com/forums/showthread.php?t=778363

this rhino related post post on RCG just came up & is generating interest, thankfully a couple people here on TS forum made the connection & got me hooked up to TS
http://www.rcgroups.com/forums/showthread.php?t=1113526

as I mentioned in the RCG post, there are fellows like this
http://www.rcuniverse.com/forum/m_7553517/tm.htm
doing high quality 3D cad work for RC scale applications & resultant construction. In this case, using a different cad program, but from what I can see of every cad screenshot & my self-taught rhino experience, nothing outside of rhino’s capabilities. His software selection just happened to originate from his particular employment background. And now with TS adding another powerful level of tools to rhino as evidenced by some recent posts, I’m really thankful I selected this software combo.

Hope this helps, good luck with your venture.

[aeronautica]

I have been working on the model in t-splines and this is what I have achieved so far with my limited knowledge of the software.

The areas where the front and rear windshield meet the fuselage are causing me problems.

Should I add points or edges to get the sharp turn between the fuselage and window?
The rear of the fuselage should be squared off. How should I do this?

The sides of the fuselage between the front and rear of the doors should be a straight vertical from the bottom of the door up. How would one achieve this?

I have done this with the box modeling mode and tried to follow the dolphin tutorials to get this far. If anyone wants to play around with the model feel free to do so. I just want to see how you do things.

Right know I am only interested in the fuselage so I can make cuts and extract formers.
On level 5 I have added lines and used the project to surface command to get some sample formers. I then export these back into Microstation to do the interior detail work for my construction plans.

Thanks

[sgreenawalt]

Hello Aeronautica -

I actually used to run a small electric rc scale kit company (now defunct, was called Acme Aircraft Co) and used AutoCAD and laser cutting for all my kits. (as an aside...need a laser cutter? ) Since leaving the kit business, I have moved onto full size experimental aviation, and I transitioned from AutoCAD to Rhino. I've spent many many hours staring at 2D drawings that look precisely what you posted in your first post. I'm going to throw out some ideas based on my experience, perhaps you might find them useful.

1. Try to keep everything in one application. Things start to get pretty spread out when you start sending things back and forth between 2D applications Rhino. Recognize that Rhino probably does all the 2D work you are used to in Microstation. Furthermore - not knowing Microstations abilities - when it comes to curves (not arcs) I found the control that I have in Rhino over curves to be far better than what I can get from an AutoCAD spline. Especially when it comes to fuselage cross sections, I will never go back to drawing them in AutoCAD, Rhino is just so much better for this task.

2. I think you might be over-thinking with respect to your fuselage. I work at the airport, so I just went out and stared at a Cessna 152 for a few minutes. There's really only four parts on the entire fuselage that you truly need 3D modeling for. They are:
1. The bottom of the cowling
2. The nose bowl
3. The windshield
4. The rear window
Everything else on the fuselage is made by forming flat sheet metal over formers - which means there is no compound curvature. One thing that always bugged me about scale models is that they were often made "prettier/smoother/curvier" than the real thing. Often times sheet metal cannot do the things that scale modelers say it can, well not without forming dies and such. Take the tailcone for example - well just the section between the back of the rear window and the leading edge of the stab. You can tell just by seeing the three vertical rows of rivets that there's three bulkheads in that area. If you have the shape of the bulkheads, and their distances/orientations to each other, you have everything you need to know about that surface. Nothing else can happen in that area, because the sheet metal won't allow it. So what I would do is to create those cross sections accurately in Rhino, and then simply use the Loft command to make that particular section. Repeat that process for the rest of the plane, and you'll have 95% of it done. The remaining bits, the ones listed above can either be done in Rhino or using Tsplines, depending on what's easiest for you.

3. Most 3-views are ok at best, if not outright terrible. If your goal is to make museum quality planes, there's no substitute to taking measurements off the real thing. I don't know where you are, but I can guarantee there's a Cessna 152 close by. Cook up some sort of rig for tracing the cross sections of aircraft (I've used multiple t-squares, cardboard with the pen/washer trick, whatever works), get those cross sections scanned in at your local blueprint shop and then make your model from that. If you're anywhere in Northern CA and ask nicely, I might even let you come and use my digitizing arm. It works directly with Rhino btw.....that whole "keep everything in one application" mantra.

[SDS]

SGreenawalt: Good points all, there's nothing like the voice of experience. We could have quite a discussion comparing "real" to CAD models.
Aeronautica: Did you draw those three views, line by line, or use some software conversion?
Regarding your TSpline modeling questions, it is easier for me to start with your three views and redraw the model. Everybody develops their own techniques. I prefer TSLines and use box modeling only for basic shapes. Review the attached files and post again!

[aeronautica]

SDS,

I develop my 3-views from as many that I can find of a subject aircraft and also photos. I then import them as raster files into Microstation and trace over them to create my own 3-views. I do my best to have the top, side and front match each other as I fine most of the old hand drawn 3-views do not match up most of the time.

I have been only trying to do the box method and have not yet tried from lines. I am reaching my current knowledge limits of the software and I am not sure I follow what you did in the samples you posted.

I am completely convinced though that Rhino and tsplines are the way to achieve what I am after and once I can master the software and develop a system for using it I can create the plans and laser cut kits that I want in a short amount of time.

Are you available for any specialized training over the internet? I have several aircraft that I want to develop into kits in as short amount of time as possible. I am in the eastern time zone.

Larry Folk / 614-604-9859
Thanks

[aeronautica]

Here is my lastest attemp so far and a picture of the real thing.
Comments are welcome.

Larry

[ricardo]

Hi Aeronautica,

See attached, I moved down the cockpit upper panel you designed and extruded the faces up to recreate it. This is really rough but gives you an idea on how to isolate the cockpit from the fuselage to get better/independent control of both.

Ricardo

[SDS]

Open "Cessna 152 Step 2", Select all the lines, Select "From Lines", then the "Face Layout" button. Select all 4 red end faces (Simple Edge Loops) to remove these, then Enter and "OK". You should see a TSpline surface which looks familiar. I started with the center section because the wing/fuselage interface is the most difficult part of this model. Attached is a file showing the results of further point editing to add more curvature to the windshield and rear window areas.
I'll offer what help I can through this Forum, but I'm not looking for a job (I'm very happily retired.)

[ricardo]

Open "Cessna 152 Step 2", Select all the lines, Select "From Lines", then the "Face Layout" button. Select all 4 red end faces (Simple Edge Loops) to remove these, then Enter and "OK". You should see a TSpline surface which looks familiar. I started with the center section because the wing/fuselage interface is the most difficult part of this model. Attached is a file showing the results of further point editing to add more curvature to the windshield and rear window areas.
I'll offer what help I can through this Forum, but I'm not looking for a job (I'm very happily retired.)

I'm glad you went with the extruded cockpit also, I'm kinda rusty and wasn't really sure that was the absolutely correct choice.

[ernchesto]

SDS - just want to thank your for posting that wireframe. I'm playing with and learning lots from it.
Cheers!

[SDS]

Aeronautica: I "finished " the Cessna, examine this and see if it's usable for your model building. The next step is to "Nurberize" the model and start adding details. The wing struts, landing gear and horizontal stabilizer are separate surfaces; I could not get a satisfactory blend for the strut fairings where they join the fuselage thus these intersections are sloppy. It's attached in box mode.
Ricardo: The only extrusion I used was to draw the basic fuselage "tub". I'm still stuck using "TS From Lines" similar to the model in the Wing Fillet thread.
Ernchesto: Thanks, we're all here to learn.

[aeronautica]

SDS,

Thanks for doing this. I am going to have to start trying the ts from lines method.
I think that I am going to import this model into a new file, lock it and try to duplicate what you did.

Thanks again.

Larry

[JSantocono]

Nice model!
Good job.

JuanS

[SDS]

Here's a couple of renders of our Cessna project after adding a few details. I'm new to rendering and am not an artist, but the software (Flamingo Nxt) does make a difference!

[ricardo]

Ricardo: The only extrusion I used was to draw the basic fuselage "tub". I'm still stuck using "TS From Lines" similar to the model in the Wing Fillet thread.

OK, I wasn't clear. I was referring to the star points on the four corners of the cockpit. I got them by extruding the top panels of the tube on the original fuselage. You from a wire cage set that way.

Great work BTW.

[JSantocono]

NIce job SDS!!

JuanS

[Matt Sederberg]

Very nice.

[cheeky]

WoW SDS, this is way cool.
I have to design a gyro body and am battling with Rhino to get what I want.
After the design I have to cut a plug with my 3d router.
Whish I could be where you are.

This is a real gyro (with the mechanics already made and flying, but they are waiting for me for the body.It looks so drab and boring. Not futuristic at all.

Peter,
South Africa

[SDS]

Sounds like a fun project; is this a gyro-copter, like a Benson?

[cheeky]

Hi SDS.
Yes, you can say like a Benson only a lot more advanced.
As I said I have to design the "looks" and mine looks drab.
Maybee I should throw it out there as a "Prizelesscompetition" to those wizzkids and their keyboard paintbrushes?

The project is quote under wraps right now so I really cannot reveal which company I am doing this for but I can post a pic of my drab look here.
I need something sexy and appealing. The shape has very few design constraints.
It must be convertable from open to closed cockpit without leaving the doors where you started your flight and it must be easy to get into.
Structural restrictions are the height of the engine from the ground/ boom (if it has one )
This model HAS to be a pusher.
[img]C:\WAGtail\BashMi\single.bmp[/img]