Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Show off your FreeCAD projects here!
User avatar
ppemawm
Posts: 447
Joined: Fri May 17, 2013 3:54 pm
Location: Manhattan New York

Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Postby ppemawm » Sat Apr 28, 2018 4:46 pm

In a previous project (https://forum.freecadweb.org/viewtopic. ... 04#p217871) I used a master sketch to link all the reciprocating components. In this much simpler project I wanted to explore any advantages of using chained body shapebinders to link the reciprocating components rather than a master sketch. Body shapebinders provide more 3D context when creating the model IMHO.

This model was created from public domain drawings of a 2009 design copyrighted by G.A. Graham called the 'Lobo Pup Twin'. Several screenshots of the the work process are included with a few comments:

Graham_twin-aero-motor.JPG
The model tree is organized as sub-assembly parts and individual bodies. The origin is located to take advantage of the obvious symmetry. In that way many of the bodies could be mirrored using the Part workbench which dramatically reduces modeling time.

Most of the sketches were attached to the origin planes in as much as possible to maximize feature stability. A few sketches were attached to body shapebinders to simplify locating components. No datum planes were used.
Graham_twin-aero-motor.JPG (227.93 KiB) Viewed 2005 times
Block&Internals.jpg
As you can see in the model tree, PartWB>mirrors across the YZ plane were used to duplicate the front bodies from the rear bodies. This kicks the mirror out of the Part subassembly but can be easily returned with drag and drop onto the proper PD Part in the tree.

ScrewMaker 2.0 fasteners were copied (CTRL C) from a V0.16 file, pasted into V0.17 (CTRL V) and arranged using Part>polar arrays located with Placement properties. The ScrewMaker macro does not produce valid solids in V0.17 in my experience.

I tried using the V0.17 Fasteners workbench but find it awkward to adjust screw lengths. However, that may just be me not willing to completely give up V0.16 habits yet.
Block&Internals.jpg (201.28 KiB) Viewed 2005 times
crank_assembly1.jpg
The crank sub-assembly began with a revolve and a sketch of a journal Pad feature on the center crankshaft which controls the crank angle property as shown in the figure. The rear rotating or reciprocating bodies are then 'chain' linked with shapebinders of each mating body. Each new body sketch is attached to its mating shapebinder. If the Sketch058 crank_angle constraint is changed, each shapebinder and its attached body reliably updates to its new location.

A master sketch was also used for the front connecting rod, piston, and wrist pin to study the effects of linking sketches to chained shapebinders. This seemed to work just as expected.
crank_assembly1.jpg (172.61 KiB) Viewed 2005 times
crank_assembly2.jpg
This figure shows the model updating to different crank angles when changed in the Sketch contraints list in the property panel of the combo view. The model proved to be quite stable but randomly failed at times due to the sketcher solver providing an alternate solution to my chosen constraints. Once can learn much about proper contraint methodology by subjecting your sketches to such a wide variation of changes. Still learning.
crank_assembly2.jpg (192.6 KiB) Viewed 2005 times
Exhaust.jpg
The exhaust was created first as a solid pipe from revolves and pads of faces along with an additive loft for the exhaust exit pipe. The PartDesign>Make a solid (thickness) tool was used to hollow out the exhaust without any difficulties.

The front pipe pad was located by a sketch attachment to the block shapebinder followed by a 90 deg revolve with a second pad to the rear junction. A reverse revolve at the junction was used to create the rear pipe ahead of the exit pipe loft. Datum lines provided convenient revolve axes.
Exhaust.jpg (141.99 KiB) Viewed 2005 times

I do prefer using chained linked body shapebinders rather than linking back to a single master sketch but have yet to fully understand if the model is as stable. One disadvantage is that if any of the shapebinders features are changed then it is likely that any dependent body will blow-up. I plan to investigate the master sketch approach once again with a more complex beam engine model currently in process.

The 7.6 Mb file is too large to attach but I can put it on Dropbox if anyone is interested in the details.

OS: Windows 10
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.17.13509 (Git)
Build type: Release
Branch: releases/FreeCAD-0-17
Hash: 0258808ccb6ba3bd5ea9312f79cd023f1a8671b7
Python version: 2.7.14
Qt version: 4.8.7
Coin version: 4.0.0a
OCC version: 7.2.0
Locale: English/UnitedStates (en_US)
chrisb
Posts: 10913
Joined: Tue Mar 17, 2015 9:14 am

Re: Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Postby chrisb » Sat Apr 28, 2018 5:27 pm

What a great model! If you you would put it in your dropbox I would like to see how you work with the ShapeBinders.
thschrader
Posts: 1171
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Postby thschrader » Sat Apr 28, 2018 5:35 pm

ppemawm wrote:
Sat Apr 28, 2018 4:46 pm
In this much simpler project...
:D :D :D
Yes, please give us a dropbox link.
User avatar
ppemawm
Posts: 447
Joined: Fri May 17, 2013 3:54 pm
Location: Manhattan New York

Re: Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Postby ppemawm » Sat Apr 28, 2018 6:49 pm

chrisb wrote:
Sat Apr 28, 2018 5:27 pm
What a great model! If you you would put it in your dropbox I would like to see how you work with the ShapeBinders.
thschrader wrote:
Sat Apr 28, 2018 5:35 pm
Yes, please give us a dropbox link.
https://www.dropbox.com/s/6d4knu40dnaw2 ... FCStd?dl=0
Questions and comments are welcome.
chrisb
Posts: 10913
Joined: Tue Mar 17, 2015 9:14 am

Re: Twin Aeromotor--V0.17 Part Design--Chained Shapebinders

Postby chrisb » Sat Apr 28, 2018 7:21 pm

ppemawm wrote:
Sat Apr 28, 2018 6:49 pm
https://www.dropbox.com/s/6d4knu40dnaw2 ... FCStd?dl=0
Thank you very much. There is much to learn from your model. Starting from where you place the center of the models to the colour finish.
It will take some time to study it.